![ing maintain the cables attenuation characteristics. The coating also helps protect the fiber
from moisture, which reduces the possibility of the occurrence of a detrimental phenome-
non called stress corrosion (sometimes called static fatigue) caused by high humidity.
Moisture causes silicon dioxide crystals to interact, causing bonds to break down and spon-
taneous fractures to form over a prolonged period of time. The protective coating is sur-
rounded by a buffer jacket, which provides the cable additional protection against abrasion
and shock. Materials commonly used for the buffer jacket include steel, fiberglass, plastic,
flame-retardant polyvinyl chloride (FR-PVC), Kevlar yarn, and paper. The buffer jacket is
encapsulated in a strength member, which increases the tensile strength of the overall cable
assembly. Finally, the entire cable assembly is contained in an outer polyurethane jacket.
There are three essential types of optical fibers commonly used today. All three vari-
eties are constructed of either glass, plastic, or a combination of glass and plastic:
Plastic core and cladding
Glass core with plastic cladding (called PCS fiber [plastic-clad silica])
Glass core and glass cladding (called SCS [silica-clad silica])
Plastic fibers are more flexible and, consequently, more rugged than glass. Therefore,
plastic cables are easier to install, can better withstand stress, are less expensive, and weigh
approximately 60% less than glass. However, plastic fibers have higher attenuation charac-
teristics and do not propagate light as efficiently as glass. Therefore, plastic fibers are lim-
ited to relatively short cable runs, such as within a single building.
Fibers with glass cores have less attenuation than plastic fibers, with PCS being
slightly better than SCS. PCS fibers are also less affected by radiation and, therefore, are
more immune to external interference. SCS fibers have the best propagation characteristics
and are easier to terminate than PCS fibers. Unfortunately, SCS fibers are the least rugged,
and they are more susceptible to increases in attenuation when exposed to radiation.
The selection of a fiber for a given application is a function of the specific system re-
quirements. There are always trade-offs based on the economics and logistics of a particu-
lar application.
6-1-1 Cable configurations. There are many different cable designs available today.
Figure 5 shows examples of several optical fiber cable configurations. With loose tube con-
struction (Figure 5a), each fiber is contained in a protective tube. Inside the tube, a
polyurethane compound encapsules the fiber and prevents the intrusion of water. A phe-
nomenon called stress corrosion or static fatigue can result if the glass fiber is exposed to
long periods of high humidity. Silicon dioxide crystals interact with the moisture and cause
bonds to break down, causing spontaneous fractures to form over a prolonged period. Some
fiber cables have more than one protective coating to ensure that the fiber’s characteristics
do not alter if the fiber is exposed to extreme temperature changes. Surrounding the fiber’s
cladding is usually a coating of either lacquer, silicon, or acrylate that is typically applied
to seal and preserve the fiber’s strength and attenuation characteristics.
Figure 5b shows the construction of a constrained optical fiber cable. Surrounding the
fiber are a primary and a secondary buffer comprised of Kevlar yarn, which increases the
tensile strength of the cable and provides protection from external mechanical influences
that could cause fiber breakage or excessive optical attenuation. Again, an outer protective
tube is filled with polyurethane, which prevents moisture from coming into contact with the
fiber core.
Figure 5c shows a multiple-strand cable configuration, which includes a steel central
member and a layer of Mylar tape wrap to increase the cable’s tensile strength. Figure 5d
shows a ribbon configuration for a telephone cable, and Figure 5e shows both end and side
views of a PCS cable.
Optical Fiber Transmission Media
8](https://image.slidesharecdn.com/advancedelectroniccommunicationssystems-221223061154-405db6b3/85/Advanced_Electronic_Communications_Systems-pdf-13-320.jpg)
![Example 3
For a single-mode optical cable with 0.25-dB/km loss, determine the optical power 100 km from a
0.1-mW light source.
Solution Substituting into Equation 15 gives
P 0.1mW 10{[(0.25)(100)]/(10)}
1 104
10{[(0.25)(100)]/(10)}
(1 104
)(1 102.5
)
0.316 μW
and
or by substituting into Equation 16
Transmission losses in optical fiber cables are one of the most important characteristics of
the fibers. Losses in the fiber result in a reduction in the light power, thus reducing the sys-
tem bandwidth, information transmission rate, efficiency, and overall system capacity. The
predominant losses in optical fiber cables are the following:
Absorption loss
Material, or Rayleigh, scattering losses
Chromatic, or wavelength, dispersion
Radiation losses
Modal dispersion
Coupling losses
10-1 Absorption Losses
Absorption losses in optical fibers is analogous to power dissipation in copper cables; im-
purities in the fiber absorb the light and convert it to heat. The ultrapure glass used to man-
ufacture optical fibers is approximately 99.9999% pure. Still, absorption losses between 1
dB/km and 1000 dB/km are typical. Essentially, there are three factors that contribute to the
absorption losses in optical fibers: ultraviolet absorption, infrared absorption, and ion res-
onance absorption.
10-1-1 Ultraviolet absorption. Ultraviolet absorption is caused by valence elec-
trons in the silica material from which fibers are manufactured. Light ionizes the valence
electrons into conduction. The ionization is equivalent to a loss in the total light field and,
consequently, contributes to the transmission losses of the fiber.
10-1-2 Infrared absorption. Infrared absorption is a result of photons of light that
are absorbed by the atoms of the glass core molecules. The absorbed photons are converted
to random mechanical vibrations typical of heating.
10-1-3 Ion resonance absorption. Ion resonance absorption is caused by OH
ions in the material. The source of the OH
ions is water molecules that have been trapped
in the glass during the manufacturing process. Iron, copper, and chromium molecules also
cause ion absorption.
35 dBm
10 dBm 25 dB
P1dBm2 10 log¢
0.1 mW
0.001 W
≤ 31100 km210.25 dBkm2 4
35 dBm
P1dBm2 10 log¢
0.316 μW
0.001
≤
Optical Fiber Transmission Media
25](https://image.slidesharecdn.com/advancedelectroniccommunicationssystems-221223061154-405db6b3/85/Advanced_Electronic_Communications_Systems-pdf-30-320.jpg)
![Optical Fiber Transmission Media
FIGURE 29 Planar heterojunction
LED
(0.5 mW [3 dBm] at 20 mA to 3.4 mW [5.3 dBm] at 140 mA). Figure 30b shows output
power versus temperature. It can be seen that the output power varies inversely with tem-
perature between a temperature range of 40°C to 80°C. Figure 30c shows relative output
power in respect to output wavelength. For this particular example, the maximum output
power is achieved at an output wavelength of 825 nm.
12-4 Burrus Etched-Well Surface-Emitting LED
For the more practical applications, such as telecommunications, data rates in excess of 100
Mbps are required. For these applications, the etched-well LED was developed. Burrus and
Dawson of Bell Laboratories developed the etched-well LED. It is a surface-emitting LED
and is shown in Figure 31. The Burrus etched-well LED emits light in many directions. The
etched well helps concentrate the emitted light to a very small area. Also, domed lenses can
be placed over the emitting surface to direct the light into a smaller area. These devices are
more efficient than the standard surface emitters, and they allow more power to be coupled
into the optical fiber, but they are also more difficult and expensive to manufacture.
12-5 Edge-Emitting LED
The edge-emitting LED, which was developed by RCA, is shown in Figure 32. These LEDs
emit a more directional light pattern than do the surface-emitting LEDs. The construction
is similar to the planar and Burrus diodes except that the emitting surface is a stripe rather
than a confined circular area. The light is emitted from an active stripe and forms an ellip-
tical beam. Surface-emitting LEDs are more commonly used than edge emitters because
they emit more light. However, the coupling losses with surface emitters are greater, and
they have narrower bandwidths.
The radiant light power emitted from an LED is a linear function of the forward cur-
rent passing through the device (Figure 33). It can also be seen that the optical output power
of an LED is, in part, a function of the operating temperature.
12-6 ILD
Lasers are constructed from many different materials, including gases, liquids, and solids,
although the type of laser used most often for fiber-optic communications is the semicon-
ductor laser.
The ILD is similar to the LED. In fact, below a certain threshold current, an ILD acts
similarly to an LED. Above the threshold current, an ILD oscillates; lasing occurs. As cur-
rent passes through a forward-biased p-n junction diode, light is emitted by spontaneous
emission at a frequency determined by the energy gap of the semiconductor material. When
a particular current level is reached, the number of minority carriers and photons produced
on either side of the p-n junction reaches a level where they begin to collide with already
excited minority carriers. This causes an increase in the ionization energy level and makes
the carriers unstable. When this happens, a typical carrier recombines with an opposite type
35](https://image.slidesharecdn.com/advancedelectroniccommunicationssystems-221223061154-405db6b3/85/Advanced_Electronic_Communications_Systems-pdf-40-320.jpg)
![If information bits are encoded (grouped) and then converted to signals with more
than two levels, transmission rates in excess of 2B are possible, as will be seen in subse-
quent sections of this chapter. In addition, since baud is the encoded rate of change, it also
equals the bit rate divided by the number of bits encoded into one signaling element. Thus,
(11)
By comparing Equation 10 with Equation 11, it can be seen that with digital modu-
lation, the baud and the ideal minimum Nyquist bandwidth have the same value and are
equal to the bit rate divided by the number of bits encoded. This statement holds true for all
forms of digital modulation except frequency-shift keying.
3 AMPLITUDE-SHIFT KEYING
The simplest digital modulation technique is amplitude-shift keying (ASK), where a binary
information signal directly modulates the amplitude of an analog carrier. ASK is similar
to standard amplitude modulation except there are only two output amplitudes possible.
Amplitude-shift keying is sometimes called digital amplitude modulation (DAM). Mathe-
matically, amplitude-shift keying is
(12)
where vask(t) amplitude-shift keying wave
vm(t) digital information (modulating) signal (volts)
A/2 unmodulated carrier amplitude (volts)
ωc analog carrier radian frequency (radians per second, 2π fct)
In Equation 12, the modulating signal (vm[t]) is a normalized binary waveform, where 1
V logic 1 and 1 V logic 0. Therefore, for a logic 1 input, vm(t) 1 V, Equation 12 re-
duces to
and for a logic 0 input, vm(t) 1 V, Equation 12 reduces to
0
Thus, the modulated wave vask(t), is either A cos(ωct) or 0. Hence, the carrier is either “on” or
“off,” which is why amplitude-shift keying is sometimes referred to as on-off keying (OOK).
Figure 2 shows the input and output waveforms from an ASK modulator. From the
figure, it can be seen that for every change in the input binary data stream, there is one
change in the ASK waveform, and the time of one bit (tb) equals the time of one analog sig-
naling element (ts). It is also important to note that for the entire time the binary input is high,
the output is a constant-amplitude, constant-frequency signal, and for the entire time the bi-
nary input is low, the carrier is off. The bit time is the reciprocal of the bit rate and the time
of one signaling element is the reciprocal of the baud. Therefore, the rate of change of the
v1ask21t2 31 14B
A
2
cos1ct2R
A cos1ct2
v1ask21t2 31 14B
A
2
cos1ct2R
v1ask21t2 31 vm1t2 4B
A
2
cos1ct2R
baud
fb
N
Digital Modulation
54](https://image.slidesharecdn.com/advancedelectroniccommunicationssystems-221223061154-405db6b3/85/Advanced_Electronic_Communications_Systems-pdf-59-320.jpg)
![(a)
Binary
input
DAM
output
(b)
FIGURE 2 Digital amplitude modula-
tion: (a) input binary; (b) output DAM
waveform
ASK waveform (baud) is the same as the rate of change of the binary input (bps); thus, the
bit rate equals the baud. With ASK, the bit rate is also equal to the minimum Nyquist band-
width. This can be verified by substituting into Equations 10 and 11 and setting N to 1:
Example 1
Determine the baud and minimum bandwidth necessary to pass a 10 kbps binary signal using ampli-
tude shift keying.
Solution For ASK, N 1, and the baud and minimum bandwidth are determined from Equations
11 and 10, respectively:
The use of amplitude-modulated analog carriers to transport digital information is a relatively
low-quality, low-cost type of digital modulation and, therefore, is seldom used except for very low-
speed telemetry circuits.
4 FREQUENCY-SHIFT KEYING
Frequency-shift keying (FSK) is another relatively simple, low-performance type of digital
modulation. FSK is a form of constant-amplitude angle modulation similar to standard fre-
quency modulation (FM) except the modulating signal is a binary signal that varies between
two discrete voltage levels rather than a continuously changing analog waveform. Conse-
quently, FSK is sometimes called binary FSK (BFSK). The general expression for FSK is
vfsk(t) Vc cos{2π[fc vm(t) f]t} (13)
where vfsk(t) binary FSK waveform
Vc peak analog carrier amplitude (volts)
fc analog carrier center frequency (hertz)
f peak change (shift) in the analog carrier frequency (hertz)
vm(t) binary input (modulating) signal (volts)
From Equation 13, it can be seen that the peak shift in the carrier frequency (f) is
proportional to the amplitude of the binary input signal (vm[t]), and the direction of the shift
baud
10,000
1
10,000
B
10,000
1
10,000
B
fb
1
fb baud
fb
1
fb
Digital Modulation
55](https://image.slidesharecdn.com/advancedelectroniccommunicationssystems-221223061154-405db6b3/85/Advanced_Electronic_Communications_Systems-pdf-60-320.jpg)
![–Δƒ +Δƒ
ƒc
ƒs ƒm
Binary input
signal
Logic 1
Logic 0
FIGURE 3 FSK in the frequency
domain
is determined by the polarity. The modulating signal is a normalized binary waveform
where a logic 1 1 V and a logic 0 1 V. Thus, for a logic 1 input, vm(t) 1, Equa-
tion 13 can be rewritten as
vfsk(t) Vc cos[2π(fc f)t]
For a logic 0 input, vm(t) 1, Equation 13 becomes
vfsk(t) Vc cos[2π(fc f)t]
With binary FSK, the carrier center frequency (fc) is shifted (deviated) up and down
in the frequency domain by the binary input signal as shown in Figure 3. As the binary in-
put signal changes from a logic 0 to a logic 1 and vice versa, the output frequency shifts be-
tween two frequencies: a mark, or logic 1 frequency (fm), and a space, or logic 0 frequency
(fs). The mark and space frequencies are separated from the carrier frequency by the peak
frequency deviation (f) and from each other by 2 f.
With FSK, frequency deviation is defined as the difference between either the mark
or space frequency and the center frequency, or half the difference between the mark and
space frequencies. Frequency deviation is illustrated in Figure 3 and expressed mathemat-
ically as
(14)
where f frequency deviation (hertz)
|fm fs| absolute difference between the mark and space frequencies (hertz)
Figure 4a shows in the time domain the binary input to an FSK modulator and the cor-
responding FSK output.As the figure shows, when the binary input (fb) changes from a logic
1 to a logic 0 and vice versa, the FSK output frequency shifts from a mark (fm) to a space (fs)
frequency and vice versa. In Figure 4a, the mark frequency is the higher frequency (fc
f), and the space frequency is the lower frequency (fc f), although this relationship could
be just the opposite. Figure 4b shows the truth table for a binary FSK modulator. The truth
table shows the input and output possibilities for a given digital modulation scheme.
4-1 FSK Bit Rate, Baud, and Bandwidth
In Figure 4a, it can be seen that the time of one bit (tb) is the same as the time the FSK out-
put is a mark of space frequency (ts). Thus, the bit time equals the time of an FSK signal-
ing element, and the bit rate equals the baud.
¢f
冟 fm fs冟
2
Digital Modulation
56](https://image.slidesharecdn.com/advancedelectroniccommunicationssystems-221223061154-405db6b3/85/Advanced_Electronic_Communications_Systems-pdf-61-320.jpg)
![FIGURE 9 PLL-FSK demodulator
FIGURE 10 Noncontinuous FSK
waveform
The most common circuit used for demodulating binary FSK signals is the phase-
locked loop (PLL), which is shown in block diagram form in Figure 9. A PLL-FSK de-
modulator works similarly to a PLL-FM demodulator. As the input to the PLL shifts be-
tween the mark and space frequencies, the dc error voltage at the output of the phase
comparator follows the frequency shift. Because there are only two input frequencies (mark
and space), there are also only two output error voltages. One represents a logic 1 and the
other a logic 0. Therefore, the output is a two-level (binary) representation of the FSK in-
put. Generally, the natural frequency of the PLL is made equal to the center frequency of
the FSK modulator. As a result, the changes in the dc error voltage follow the changes in
the analog input frequency and are symmetrical around 0 V.
Binary FSK has a poorer error performance than PSK or QAM and, consequently, is sel-
dom used for high-performance digital radio systems. Its use is restricted to low-performance,
low-cost, asynchronous data modems that are used for data communications over analog,
voice-band telephone lines.
4-4 Continuous-Phase Frequency-Shift Keying
Continuous-phase frequency-shift keying (CP-FSK) is binary FSK except the mark and
space frequencies are synchronized with the input binary bit rate. Synchronous simply im-
plies that there is a precise time relationship between the two; it does not mean they are equal.
With CP-FSK, the mark and space frequencies are selected such that they are separated from
the center frequency by an exact multiple of one-half the bit rate (fm and fs n[fb /2]), where
n any integer). This ensures a smooth phase transition in the analog output signal when it
changes from a mark to a space frequency or vice versa. Figure 10 shows a noncontinuous
FSK waveform. It can be seen that when the input changes from a logic 1 to a logic 0 and
vice versa, there is an abrupt phase discontinuity in the analog signal. When this occurs, the
demodulator has trouble following the frequency shift; consequently, an error may occur.
Figure 11 shows a continuous phase FSK waveform. Notice that when the output fre-
quency changes, it is a smooth, continuous transition. Consequently, there are no phase dis-
continuities. CP-FSK has a better bit-error performance than conventional binary FSK for
a given signal-to-noise ratio. The disadvantage of CP-FSK is that it requires synchroniza-
tion circuits and is, therefore, more expensive to implement.
Digital Modulation
61](https://image.slidesharecdn.com/advancedelectroniccommunicationssystems-221223061154-405db6b3/85/Advanced_Electronic_Communications_Systems-pdf-66-320.jpg)
![FIGURE 14 BPSK modulator: (a) truth table; (b) phasor
diagram; (c) constellation diagram
transformer T2 in phase with the carrier voltage across T1. Consequently, the output signal
is in phase with the reference oscillator.
If the binary input is a logic 0 (negative voltage), diodes D1 and D2 are reverse biased
and off, while diodes D3 and D4 are forward biased and on (Figure 13c).As a result, the car-
rier voltage is developed across transformer T2 180° out of phase with the carrier voltage
across T1. Consequently, the output signal is 180° out of phase with the reference oscillator.
Figure 14 shows the truth table, phasor diagram, and constellation diagram for a BPSK mod-
ulator. A constellation diagram, which is sometimes called a signal state-space diagram, is
similar to a phasor diagram except that the entire phasor is not drawn. In a constellation di-
agram, only the relative positions of the peaks of the phasors are shown.
5-1-2 Bandwidth considerations of BPSK. A balanced modulator is a product
modulator; the output signal is the product of the two input signals. In a BPSK modulator,
the carrier input signal is multiplied by the binary data. If 1 V is assigned to a logic 1 and
1 V is assigned to a logic 0, the input carrier (sin ωct) is multiplied by either a or 1.
Consequently, the output signal is either 1 sin ωct or 1 sin ωct; the first represents a sig-
nal that is in phase with the reference oscillator, the latter a signal that is 180° out of phase
with the reference oscillator. Each time the input logic condition changes, the output phase
changes. Consequently, for BPSK, the output rate of change (baud) is equal to the input rate
of change (bps), and the widest output bandwidth occurs when the input binary data are an
alternating 1/0 sequence. The fundamental frequency ( fa) of an alternative 1/0 bit sequence
is equal to one-half of the bit rate ( fb/2). Mathematically, the output of a BPSK modulator
is proportional to
BPSK output [sin(2πfat)] [sin(2πfct)] (20)
Digital Modulation
64](https://image.slidesharecdn.com/advancedelectroniccommunicationssystems-221223061154-405db6b3/85/Advanced_Electronic_Communications_Systems-pdf-69-320.jpg)
![BPF LPF
Balanced
modulator
BPSK
input
± sin(ωct)
sin(ωct)
UP
Binary
data
output
Coherent
carrier
recovery
Level
converter
Clock
recovery
+v
o
FIGURE 16 Block diagram of a BPSK receiver
Solution Substituting into Equation 20 yields
output (sin ωat)(sin ωct)
[sin 2π(5 MHz)t][sin 2π(70 MHz)t]
lower side frequency upper side frequency
Minimum lower side frequency (LSF):
LSF 70 MHz 5 MHz 65 MHz
Maximum upper side frequency (USF):
USF 70 MHz 5 MHz 75 MHz
Therefore, the output spectrum for the worst-case binary input conditions is as follows:
The minimum Nyquist bandwidth (B) is
1
2
cos 2π 170 MHz 5 MHz2t
1
2
cos 2π 170 MHz 5 MHz2t
Digital Modulation
冧
冧
B 75 MHz 65 MHz 10 MHz
and the baud fb or 10 megabaud.
5-1-3 BPSK receiver. Figure 16 shows the block diagram of a BPSK receiver. The
input signal may be sin ωct or sin ωct. The coherent carrier recovery circuit detects and
regenerates a carrier signal that is both frequency and phase coherent with the original
transmit carrier. The balanced modulator is a product detector; the output is the product of
the two inputs (the BPSK signal and the recovered carrier). The low-pass filter (LPF) sep-
arates the recovered binary data from the complex demodulated signal. Mathematically, the
demodulation process is as follows.
For a BPSK input signal of sin ωct (logic 1), the output of the balanced modulator is
output (sin ωct)(sin ωct) sin2
ωct (21)
66](https://image.slidesharecdn.com/advancedelectroniccommunicationssystems-221223061154-405db6b3/85/Advanced_Electronic_Communications_Systems-pdf-71-320.jpg)
![or
leaving
It can be seen that the output of the balanced modulator contains a positive voltage
([1/2]V) and a cosine wave at twice the carrier frequency (2 ωc) The LPF has a cutoff fre-
quency much lower than 2ωc and, thus, blocks the second harmonic of the carrier and
passes only the positive constant component. A positive voltage represents a demodulated
logic 1.
For a BPSK input signal of sin ωct (logic 0), the output of the balanced modulator is
output (sin ωct)(sin ωct) sin2
ωct
(filtered out)
or
leaving
The output of the balanced modulator contains a negative voltage ([1/2]V) and a
cosine wave at twice the carrier frequency (2ωc). Again, the LPF blocks the second har-
monic of the carrier and passes only the negative constant component. A negative voltage
represents a demodulated logic 0.
5-2 Quaternary Phase-Shift Keying
Quaternary phase shift keying (QPSK), or quadrature PSK as it is sometimes called, is an-
other form of angle-modulated, constant-amplitude digital modulation. QPSK is an M-ary
encoding scheme where N 2 and M 4 (hence, the name “quaternary” meaning “4”).
With QPSK, four output phases are possible for a single carrier frequency. Because there
are four output phases, there must be four different input conditions. Because the digital in-
put to a QPSK modulator is a binary (base 2) signal, to produce four different input com-
binations, the modulator requires more than a single input bit to determine the output con-
dition. With two bits, there are four possible conditions: 00, 01, 10, and 11. Therefore, with
QPSK, the binary input data are combined into groups of two bits, called dibits. In the mod-
ulator, each dibit code generates one of the four possible output phases (45°, 135°,
45°, and 135°). Therefore, for each two-bit dibit clocked into the modulator, a single
output change occurs, and the rate of change at the output (baud) is equal to one-half the
input bit rate (i.e., two input bits produce one output phase change).
5-2-1 QPSK transmitter. A block diagram of a QPSK modulator is shown in
Figure 17. Two bits (a dibit) are clocked into the bit splitter. After both bits have been seri-
ally inputted, they are simultaneously parallel outputted. One bit is directed to the I chan-
nel and the other to the Q channel. The I bit modulates a carrier that is in phase with the ref-
erence oscillator (hence the name “I” for “in phase” channel), and the Q bit modulates a
carrier that is 90° out of phase or in quadrature with the reference carrier (hence the name
“Q” for “quadrature” channel).
It can be seen that once a dibit has been split into the I and Q channels, the operation
is the same as in a BPSK modulator. Essentially, a QPSK modulator is two BPSK modula-
tors combined in parallel. Again, for a logic 1 1 V and a logic 0 1 V, two phases
are possible at the output of the I balanced modulator (sin ωct and sin ωct), and two
output
1
2
V logic 0
sin2
ct
1
2
11 cos 2ct2
1
2
1
2
cos 2ct
output
1
2
V logic 1
sin2
ct
1
2
11 cos 2ct2
1
2
1
2
cos 2 ct
Digital Modulation
(filtered out)
67](https://image.slidesharecdn.com/advancedelectroniccommunicationssystems-221223061154-405db6b3/85/Advanced_Electronic_Communications_Systems-pdf-72-320.jpg)
![Fundamental Concepts of Data Communications
Table 2 ASCII-77: Odd Parity
Binary Code Binary Code
Bit 7 6 5 4 3 2 1 0 Hex Bit 7 6 5 4 3 2 1 0 Hex
NUL 1 0 0 0 0 0 0 0 00 @ 0 1 0 0 0 0 0 0 40
SOH 0 0 0 0 0 0 0 1 01 A 1 1 0 0 0 0 0 1 41
STX 0 0 0 0 0 0 1 0 02 B 1 1 0 0 0 0 1 0 42
ETX 1 0 0 0 0 0 1 1 03 C 0 1 0 0 0 0 1 1 43
EOT 0 0 0 0 0 1 0 0 04 D 1 1 0 0 0 1 0 0 44
ENQ 1 0 0 0 0 1 0 1 05 E 0 1 0 0 0 1 0 1 45
ACK 1 0 0 0 0 1 1 0 06 F 0 1 0 0 0 1 1 0 46
BEL 0 0 0 0 0 1 1 1 07 G 1 1 0 0 0 1 1 1 47
BS 0 0 0 0 1 0 0 0 08 H 1 1 0 0 1 0 0 0 48
HT 1 0 0 0 1 0 0 1 09 I 0 1 0 0 1 0 0 1 49
NL 1 0 0 0 1 0 1 0 0A J 0 1 0 0 1 0 1 0 4A
VT 0 0 0 0 1 0 1 1 0B K 1 1 0 0 1 0 1 1 4B
FF 1 0 0 0 1 1 0 0 0C L 0 1 0 0 1 1 0 0 4C
CR 0 0 0 0 1 1 0 1 0D M 1 1 0 0 1 1 0 1 4D
SO 0 0 0 0 1 1 1 0 0E N 1 1 0 0 1 1 1 0 4E
SI 1 0 0 0 1 1 1 1 0F O 0 1 0 0 1 1 1 1 4F
DLE 0 0 0 1 0 0 0 0 10 P 1 1 0 1 0 0 0 0 50
DC1 0 0 0 1 0 0 0 1 11 Q 0 1 0 1 0 0 0 1 51
DC2 1 0 0 1 0 0 1 0 12 R 0 1 0 1 0 0 1 0 52
DC3 0 0 0 1 0 0 1 1 13 S 1 1 0 1 0 0 1 1 53
DC4 1 0 0 1 0 1 0 0 14 T 0 1 0 1 0 1 0 0 54
NAK 0 0 0 1 0 1 0 1 15 U 1 1 0 1 0 1 0 1 55
SYN 0 0 0 1 0 1 1 0 16 V 1 1 0 1 0 1 1 0 56
ETB 1 0 0 1 0 1 1 1 17 W 0 1 0 1 0 1 1 1 57
CAN 1 0 0 1 1 0 0 0 18 X 0 1 0 1 1 0 0 0 58
EM 0 0 0 1 1 0 0 1 19 Y 1 1 0 1 1 0 0 1 59
SUB 0 0 0 1 1 0 1 0 1A Z 1 1 0 1 1 0 1 0 5A
ESC 1 0 0 1 1 0 1 1 1B [ 0 1 0 1 1 0 1 1 5B
FS 0 0 0 1 1 1 0 0 1C 1 1 0 1 1 1 0 0 5C
GS 1 0 0 1 1 1 0 1 1D ] 0 1 0 1 1 1 0 1 5D
RS 1 0 0 1 1 1 1 0 1E 0 1 0 1 1 1 1 0 5E
US 0 0 0 1 1 1 1 1 1F - 1 1 0 1 1 1 1 1 5F
SP 0 0 1 0 0 0 0 0 20 ` 1 1 1 0 0 0 0 0 60
! 1 0 1 0 0 0 0 1 21 a 0 1 1 0 0 0 0 1 61
″ 1 0 1 0 0 0 1 0 22 b 0 1 1 0 0 0 1 0 62
# 0 0 1 0 0 0 1 1 23 c 1 1 1 0 0 0 1 1 63
$ 1 0 1 0 0 1 0 0 24 d 0 1 1 0 0 1 0 0 64
% 0 0 1 0 0 1 0 1 25 e 1 1 1 0 0 1 0 1 65
0 0 1 0 0 1 1 0 26 f 1 1 1 0 0 1 1 0 66
′ 1 0 1 0 0 1 1 1 27 g 0 1 1 0 0 1 1 1 67
( 1 0 1 0 1 0 0 0 28 h 0 1 1 0 1 0 0 0 68
) 0 0 1 0 1 0 0 1 29 i 1 1 1 0 1 0 0 1 69
* 0 0 1 0 1 0 1 0 2A j 1 1 1 0 1 0 1 0 6A
1 0 1 0 1 0 1 1 2B k 0 1 1 0 1 0 1 1 6B
, 0 0 1 0 1 1 0 0 2C l 1 1 1 0 1 1 0 0 6C
- 1 0 1 0 1 1 0 1 2D m 0 1 1 0 1 1 0 1 6D
. 1 0 1 0 1 1 1 0 2E n 0 1 1 0 1 1 1 0 6E
/ 0 0 1 0 1 1 1 1 2F o 1 1 1 0 1 1 1 1 6F
0 1 0 1 1 0 0 0 0 30 p 0 1 1 1 0 0 0 0 70
1 0 0 1 1 0 0 0 1 31 q 1 1 1 1 0 0 0 1 71
2 0 0 1 1 0 0 1 0 32 r 1 1 1 1 0 0 1 0 72
3 1 0 1 1 0 0 1 1 33 s 0 1 1 1 0 0 1 1 73
4 0 0 1 1 0 1 0 0 34 t 1 1 1 1 0 1 0 0 74
5 1 0 1 1 0 1 0 1 35 u 0 1 1 1 0 1 0 1 75
6 1 0 1 1 0 1 1 0 36 v 0 1 1 1 0 1 1 0 76
7 0 0 1 1 0 1 1 1 37 w 1 1 1 1 0 1 1 1 77
8 0 0 1 1 1 0 0 0 38 x 1 1 1 1 1 0 0 0 78
(Continued)
152](https://image.slidesharecdn.com/advancedelectroniccommunicationssystems-221223061154-405db6b3/85/Advanced_Electronic_Communications_Systems-pdf-157-320.jpg)
![Error performance is the rate in which errors occur, which can be described as either
an expected or an empirical value. The theoretical (mathematical) expectation of the rate at
which errors will occur is called probability of error (P[e]), whereas the actual historical
record of a system’s error performance is called bit error rate (BER). For example, if a sys-
tem has a P(e) of 105
, this means that mathematically the system can expect to experience
one bit error for every 100,000 bits transported through the system (105
1/105
1/100,000). If a system has a BER of 105
, this means that in the past there was one bit er-
ror for every 100,000 bits transported. Typically, a BER is measured and then compared
with the probability of error to evaluate system performance. Error control can be divided
into two general categories: error detection and error correction.
5 ERROR DETECTION
Error detection is the process of monitoring data transmission and determining when errors
have occurred. Error-detection techniques neither correct errors nor identify which bits are
in error—they indicate only when an error has occurred. The purpose of error detection is
not to prevent errors from occurring but to prevent undetected errors from occurring.
The most common error-detection techniques are redundancy checking, which in-
cludes vertical redundancy checking, checksum, longitudinal redundancy checking, and
cyclic redundancy checking.
5-1 Redundancy Checking
Duplicating each data unit for the purpose of detecting errors is a form of error detection
called redundancy. Redundancy is an effective but rather costly means of detecting errors,
especially with long messages. It is much more efficient to add bits to data units that check
for transmission errors. Adding bits for the sole purpose of detecting errors is called
redundancy checking. There are four basic types of redundancy checks: vertical redundancy
checking, checksums, longitudinal redundancy checking, and cyclic redundancy checking.
5-1-1 Vertical redundancy checking. Vertical redundancy checking (VRC) is
probably the simplest error-detection scheme and is generally referred to as character par-
ity or simply parity. With character parity, each character has its own error-detection bit
called the parity bit. Since the parity bit is not actually part of the character, it is considered
a redundant bit. An n-character message would have n redundant parity bits. Therefore, the
number of error-detection bits is directly proportional to the length of the message.
With character parity, a single parity bit is added to each character to force the total
number of logic 1s in the character, including the parity bit, to be either an odd number (odd
parity) or an even number (even parity). For example, the ASCII code for the letter C is 43
hex, or P1000011 binary, where the P bit is the parity bit. There are three logic 1s in the
code, not counting the parity bit. If odd parity is used, the P bit is made a logic 0, keeping
the total number of logic 1s at three, which is an odd number. If even parity is used, the P
bit is made a logic 1, making the total number of logic 1s four, which is an even number.
The primary advantage of parity is its simplicity. The disadvantage is that when an
even number of bits are received in error, the parity checker will not detect them because
when the logic condition of an even number of bits is changed, the parity of the character
remains the same. Consequently, over a long time, parity will theoretically detect only 50%
of the transmission errors (this assumes an equal probability that an even or an odd number
of bits could be in error).
Example 2
Determine the odd and even parity bits for the ASCII character R.
Solution The hex code for the ASCII character R is 52, which is P1010010 in binary, where P des-
ignates the parity bit.
Fundamental Concepts of Data Communications
161](https://image.slidesharecdn.com/advancedelectroniccommunicationssystems-221223061154-405db6b3/85/Advanced_Electronic_Communications_Systems-pdf-166-320.jpg)
![2-2-2 Sliding window flow control. With sliding window flow control, a source
station can transmit several frames in succession before receiving an acknowledgment.
There is only one acknowledgment for several transmitted frames, thus reducing the num-
ber of acknowledgments and considerably reducing the total elapsed transmission time as
compared to stop-and-wait flow control.
The term sliding window refers to imaginary receptacles at the source and destination
stations with the capacity of holding several frames of data. Message frames can be acknowl-
edged any time before the window is filled with data. To keep track of which frames have been
acknowledged and which have not, sliding window procedures require a modulo-n number-
ing system where each transmitted frame is identified with a unique sequence number be-
tween 0 and n 1. n is any integer value equal to 2x
, where x equals the number of bits in
the numbering system. With a three-bit binary numbering system, there are 23
, or eight, pos-
sible numbers (0, 1, 2, 3, 4, 5, 6, and 7), and therefore the windows must have the capacity
of holding n 1 (seven) frames of data. The reason for limiting the number of frames to
n 1 is explained in Section 6-1-3.
The primary advantage of sliding window flow control is network utilization. With
fewer acknowledgments (i.e., fewer line turnarounds), considerably less network time is
wasted acknowledging messages, and more time can be spent actually sending messages.
The primary disadvantages of sliding window flow control are complexity and hardware
capacity. Each secondary station on a network must have sufficient buffer space to hold
2(n 1) frames of data (n 1 transmit frames and n 1 receive frames), and the primary
station must have sufficient buffer space to hold m(2[n 1]), where m equals the number of
secondary stations on the network. In addition, each secondary must store each unacknow-
ledged frame it has transmitted and keep track of the number of each unacknowledged frame
it transmits and receives. The primary station must store and keep track of all unacknow-
ledged frames it has transmitted and received for each secondary station on the network.
2-3 Error Control
Error control includes both error detection and error correction. However, with the data-link
layer, error control is concerned primarily with error detection and message retransmission,
which is the most common method of error correction.
With poll/select line disciplines, all polls, selections, and message transmissions end
with some type of end-of-transmission sequence. In addition, all messages transported from
the primary to a secondary or from a secondary to the primary are acknowledged with ACK
or NAK sequences to verify the validity of the message. An ACK means the message was
received with no transmission errors, and a NAK means there were errors in the received
message. A NAK is an automatic call for retransmission of the last message.
Error detection at the data-link layer can be accomplished with any of the methods
you've learned, such as VRC, LRC, or CRC. Error correction is generally accomplished
with a type of retransmission called automatic repeat request (ARQ) (sometimes called aut-
omatic request for retransmission). With ARQ, when a transmission error is detected, the
destination station sends a NAK back to the source station requesting retransmission of the
last message frame or frames. ARQ also calls for retransmission of missing or lost frames,
which are frames that either never reach the secondary or are damaged so severely that the
destination station does not recognize them. ARQ also calls for retransmission of frames
where the acknowledgments (either ACKs or NAKs) are lost or damaged.
There are two types of ARQ: stop-and-wait and sliding window. Stop-and-wait flow
control generally incorporates stop-and-wait ARQ, and sliding window flow control can
implement ARQ in one of two variants: go-back-n frames or selective reject (SREJ). With
go-back-n frames, the destination station tells the source station to go back n frames and re-
transmit all of them, even if all the frames did not contain errors. Go-back-n requests re-
transmission of the damaged frame plus any other frames that were transmitted after it. If
Data-Link Protocols and Data Communications Networks
219](https://image.slidesharecdn.com/advancedelectroniccommunicationssystems-221223061154-405db6b3/85/Advanced_Electronic_Communications_Systems-pdf-224-320.jpg)
![P S S E P S S S S E P
A Y Y O A Y Y P P D D N A
D N N T D N N A A A A Q D
Device address
twice
Table 1 Station and Device Addresses
Station or Station or
Device Number SPA SSA DA Device Number SPA SSA DA
0 sp sp 16 0
1 A / A 17 J 1 J
2 B S B 18 K 2 K
3 C T C 19 L 3 L
4 D U D 20 M 4 M
5 E V E 21 N 5 N
6 F W F 22 O 6 O
7 G X G 23 P 7 P
8 H Y H 24 Q 8 Q
9 I Z I 25 R 9 R
10 [ - [ 26 ] : ]
11 . , . 27 $ # $
12 % 28 * @ *
13 ( — ( 29 ) ‘ )
14 30 ; ;
15 ! ? ! 31 ^ ” ^
The character sequence for a specific poll is identical to a general poll except two device
address (DA) characters are substituted for the two quotation marks. With a specific poll,
both the station and the device address are included. Therefore, a specific poll is an invita-
tion for only one specific device at a given secondary station to transmit its message.Again,
two DA characters are transmitted for redundancy error detection.
Table 1 lists the station polling addresses, station selection addresses, and device ad-
dresses for a BSC system with a maximum of 32 stations and 32 devices.
With bisync, there are only two ways in which a secondary station can respond to a
poll: with a formatted message or with an ACK. The character sequence for a ACK is
5-1-2 BSC selection sequence. The format for a selection with BSC is
P S S E P
A Y Y O A
D N N T D
P S S E P S S S S E P
A Y Y O A Y Y S S D D N A
D N N T D N N A A A A Q D
Device address
Station selection address
Data-Link Protocols and Data Communications Networks
With BSC, there is a second polling sequence called a specific poll. The format for a
specific poll is
224](https://image.slidesharecdn.com/advancedelectroniccommunicationssystems-221223061154-405db6b3/85/Advanced_Electronic_Communications_Systems-pdf-229-320.jpg)
![5-1 Transmission Voltages and DC Component
Transmission voltages or levels can be categorized as being either unipolar (UP) or bipolar
(BP). Unipolar transmission of binary data involves the transmission of only a single nonzero
voltage level (e.g., either a positive or a negative voltage for a logic 1 and 0 V [ground] for a
logic 0). In bipolar transmission, two nonzero voltages are involved (e.g., a positive voltage
for a logic 1 and an equal-magnitude negative voltage for a logic 0 or vice versa).
Over a digital transmission line, it is more power efficient to encode binary data with
voltages that are equal in magnitude but opposite in polarity and symmetrically balanced
about 0V. For example, assuming a 1-ohm resistance and a logic 1 level of 5V and a logic
0 level of 0 V, the average power required is 12.5 W, assuming an equal probability of the
occurrence of a logic 1 or a logic 0. With a logic 1 level of 2.5 V and a logic 0 level of
2.5 V, the average power is only 6.25 W. Thus, by using bipolar symmetrical voltages, the
average power is reduced by a factor of 50%.
5-2 Duty Cycle
The duty cycle of a binary pulse can be used to categorize the type of transmission. If the bi-
nary pulse is maintained for the entire bit time, this is called nonreturn to zero (NRZ). If the ac-
tive time of the binary pulse is less than 100% of the bit time, this is called return to zero (RZ).
Unipolar and bipolar transmission voltages can be combined with either RZ or NRZ
in several ways to achieve a particular line-encoding scheme. Figure 12 shows five line-en-
coding possibilities.
In Figure 12a, there is only one nonzero voltage level (V logic 1); a zero voltage
indicates a logic 0. Also, each logic 1 condition maintains the positive voltage for the en-
tire bit time (100% duty cycle). Consequently, Figure 12a represents a unipolar nonre-
turn-to-zero signal (UPNRZ). Assuming an equal number of 1s and 0s, the average dc volt-
age of a UPNRZ waveform is equal to half the nonzero voltage (V/2).
In Figure 12b, there are two nonzero voltages (V logic 1 and V logic 0) and
a 100% duty cycle is used. Therefore, Figure 12b represents a bipolar nonreturn-to-zero sig-
nal (BPNRZ). When equal-magnitude voltages are used for logic 1s and logic 0s, and as-
suming an equal probability of logic 1s and logic 0s occurring, the average dc voltage of a
BPNRZ waveform is 0 V.
In Figure 12c, only one nonzero voltage is used, but each pulse is active for only 50%
of a bit time (tb/2). Consequently, the waveform shown in Figure 12c represents a unipolar
return-to-zero signal (UPRZ).Assuming an equal probability of 1s and 0s occurring, the av-
erage dc voltage of a UPRZ waveform is one-fourth the nonzero voltage (V/4).
Figure 12d shows a waveform where there are two nonzero voltages (V logic 1
and V logic 0). Also, each pulse is active only 50% of a bit time. Consequently, the
waveform shown in Figure 8d represents a bipolar return-to-zero (BPRZ) signal. Assuming
equal-magnitude voltages for logic 1s and logic 0s and an equal probability of 1s and 0s oc-
curring, the average dc voltage of a BPRZ waveform is 0 V.
In Figure 12e, there are again two nonzero voltage levels (V and V), but now both
polarities represent logic 1s, and 0 V represents a logic 0. This method of line encoding is
called alternate mark inversion (AMI). With AMI transmissions, successive logic 1s are in-
verted in polarity from the previous logic 1. Because return to zero is used, the encoding tech-
nique is called bipolar-return-to-zero alternate mark inversion (BPRZ-AMI). The average
dc voltage of a BPRZ-AMI waveform is approximately 0V regardless of the bit sequence.
With NRZ encoding, a long string of either logic 1s or logic 0s produces a condition
in which a receive may lose its amplitude reference for optimum discrimination between
received 1s and 0s. This condition is called dc wandering. The problem may also arise when
there is a significant imbalance in the number of 1s and 0s transmitted. Figure 13 shows
how dc wandering is produced from a long string of successive logic 1s. It can be seen that
after a long string of 1s, 1-to-0 errors are more likely than 0-to-1 errors. Similarly, long
strings of logic 0s increase the probability of 0-to-1 errors.
Digital T-Carriers and Multiplexing
339](https://image.slidesharecdn.com/advancedelectroniccommunicationssystems-221223061154-405db6b3/85/Advanced_Electronic_Communications_Systems-pdf-344-320.jpg)
![scribe the total signal power present at that point—it merely defines what the ideal level
should be. The transmission level (TL) at any point in a transmission system is the ratio in
dB of the power of a signal at that point to the power the same signal would be at a 0-dBm
transmission level point. For example, a signal at a particular point in a transmission sys-
tem measures 13 dBm. Is this good or bad? This could be answered only if it is known
what the signal strength should be at that point. TLP does just that. The reference for TLP
is 0 dBm. A 15-dBm TLP indicates that, at this specific point in the transmission system,
the signal should measure 15 dBm. Therefore, the transmission level for a signal that
measures 13 dBm at a 15-dBm point is 2 dB. A 0 TLP is a TLP where the signal
power should be 0 dBm. TLP says nothing about the actual signal level itself.
Data level point (DLP) is a parameter equivalent to TLP except TLP is used for voice
circuits, whereas DLP is used as a reference for data transmission. The DLP is always 13
dB below the voice level for the same point. If the TLP is 15 dBm, the DLP at the same
point is 28 dBm. Because a data signal is more sensitive to nonlinear distortion (harmonic
and intermodulation distortion), data signals are transmitted at a lower level than voice sig-
nals.
4-3 Units of Measurement
Common units for signal and noise power measurements in the telephone industry include
dBmO, rn, dBrn, dBrnc, dBrn 3-kHz flat, and dBrncO.
4-3-1 dBmO. dBmO is dBm referenced to a zero transmission level point (0
TLP). dBmO is a power measurement adjusted to 0 dBm that indicates what the power
would be if the signal were measured at a 0 TLP. dBmO compares the actual signal level at
a point with what that signal level should be at that point. For example, a signal measuring
17 dBm at a 16-dBm transmission level point is 1 dBmO (i.e., the signal is 1 dB be-
low what it should be, or if it were measured at a 0 TLP, it would measure 1 dBm).
4-3-2 rn (reference noise). rn is the dB value used as the reference for noise read-
ings. Reference noise equals 90 dBm or 1 pW (1 1012
W). This value was selected
for two reasons: (1) Early noise measuring sets could not accurately measure noise levels
lower than 90 dBm, and (2) noise readings are typically higher than 90 dBm, resulting
in positive dB readings in respect to reference noise.
4-3-3 dBrn. dBrn is the dB level of noise with respect to reference noise (90
dBm). dBrn is seldom used by itself since it does not specify a weighting. A noise reading
of 50 dBm equates to 40 dBrn, which is 40 dB above reference noise (50 [ 90]) 40
dBrn.
4-3-4 dBrnc. dBrnc is similar to dBrn except dBrnc is the dB value of noise with
respect to reference noise using C-message weighting. Noise measurements obtained with
a C-message filter are meaningful, as they relate the noise measured to the combined fre-
quency response of a standard telephone and the human ear.
4-3-5 dBrn 3-kHz flat. dBrn 3-kHz flat noise measurements are noise readings
taken with a filter that has a flat frequency response from 30 Hz to 3 kHz. Noise readings
taken with a 3-kHz flat filter are especially useful for detecting low-frequency noise, such
as power supply hum. dBrn 3-kHz flat readings are typically 1.5 dB higher than dBrnc read-
ings for equal noise power levels.
4-3-6 dBrncO. dBrncO is the amount of noise in dBrnc corrected to a 0 TLP. A
noise reading of 34 dBrnc at a 7-dBm TLP equates to 27 dBrncO. dBrncO relates noise
power readings (dBrnc) to a 0 TLP. This unit establishes a common reference point through-
out the transmission system.
The Telephone Circuit
413](https://image.slidesharecdn.com/advancedelectroniccommunicationssystems-221223061154-405db6b3/85/Advanced_Electronic_Communications_Systems-pdf-418-320.jpg)
![anything to the circuit, and it cannot be used to improve a circuit. It simply places higher
requirements on circuits used for high-speed data transmission. Only circuits that meet
D-type conditioning requirements can be used for high-speed data transmission. D-type
conditioning is sometimes referred to as high-performance conditioning and can be applied
to private-line data circuits in addition to either basic or C-conditioned requirements. There
are two categories for D-type conditioning: D1 and D2. Limits imposed by D1 and D2 are
virtually identical. The only difference between the two categories is the circuit arrange-
ment to which they apply. D1 conditioning specifies requirements for two-point circuits,
and D2 conditioning specifies requirements for multipoint circuits.
D-type conditioning is mandatory when the data transmission rate is 9600 bps be-
cause without D-type conditioning, it is highly unlikely that the circuit can meet the mini-
mum performance requirements guaranteed by the telephone company. When a telephone
company assigns a circuit to a subscriber for use as a 9600-bps data circuit and the circuit
does not meet the minimum requirements of D-type conditioning, a new circuit is assigned.
This is because a circuit cannot generally be upgraded to meet D-type conditioning speci-
fications by simply adding corrective devices, such as equalizers and amplifiers. Telephone
companies do not guarantee the performance of data modems operating at bit rates above
9600 bps over standard voice-grade circuits.
D-type conditioned circuits must meet the following specifications:
Signal-to-C-notched noise ratio: 28 dB
Nonlinear distortion
Signal-to-second order distortion: 35 dB
Signal-to-third order distortion: 40 dB
The signal-to-notched noise ratio requirement for standard circuits is only 24 dB, and they
have no requirements for nonlinear distortion.
Nonlinear distortion is an example of correlated noise and is produced from non-
linear amplification. When an amplifier is driven into a nonlinear operating region, the
signal is distorted, producing multiples and sums and differences (cross products) the
original signal frequencies. The noise caused by nonlinear distortion is in the form of
additional frequencies produced from nonlinear amplification of a signal. In other
words, no signal, no noise. Nonlinear distortion produces distorted waveforms that are
detrimental to digitally modulated carriers used with voice-band data modems, such as
FSK, PSK, and QAM. Two classifications of nonlinear distortion are harmonic distor-
tion (unwanted multiples of the transmitted frequencies) and intermodulation distor-
tion (cross products [sums and differences] of the transmitted frequencies, sometimes
called fluctuation noise or cross-modulation noise). Harmonic and intermodulation dis-
tortion, if of sufficient magnitude, can destroy the integrity of a data signal. The degree
of circuit nonlinearity can be measured using either harmonic or intermodulation dis-
tortion tests.
Harmonic distortion is measured by applying a single-frequency test tone to a tele-
phone channel. At the receive end, the power of the fundamental, second, and third har-
monic frequencies is measured. Harmonic distortion is classified as second, third, nth or-
der, or as total harmonic distortion. The actual amount of nonlinearity in a circuit is
determined by comparing the power of the fundamental with the combined powers of the
second and third harmonics. Harmonic distortion tests use a single-frequency (704-Hz)
source (see Figure 10); therefore, no cross-product frequencies are produced.
Although simple harmonic distortion tests provide an accurate measurement of the
nonlinear characteristics of analog telephone channel, they are inadequate for digital (T car-
rier) facilities. For this reason, a more refined method was developed that uses a multifre-
quency test-tone signal. Four test frequencies are used (see Figure 11): two designated
The Telephone Circuit
422](https://image.slidesharecdn.com/advancedelectroniccommunicationssystems-221223061154-405db6b3/85/Advanced_Electronic_Communications_Systems-pdf-427-320.jpg)
Advanced_Electronic_Communications_Systems.pdf
- 1. 9 781292 027357 ISBN 978-1-29202-735-7 Advanced Electronic Communications Systems WayneTomasi Sixth Edition Advanced Electronic Communications Systems Tomasi Sixth Edition
- 2. Pearson New International Edition Advanced Electronic Communications Systems Wayne Tomasi Sixth Edition
- 3. Pearson Education Limited Edinburgh Gate Harlow Essex CM20 2JE England and Associated Companies throughout the world Visit us on the World Wide Web at: www.pearsoned.co.uk © Pearson Education Limited 2014 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without either the prior written permission of the publisher or a licence permitting restricted copying in the United Kingdom issued by the Copyright Licensing Agency Ltd, Saffron House, 6–10 Kirby Street, London EC1N 8TS. All trademarks used herein are the property of their respective owners. The use of any trademark in this text does not vest in the author or publisher any trademark ownership rights in such trademarks, nor does the use of such trademarks imply any affiliation with or endorsement of this book by such owners. ISBN 10: 1-269-37450-8 ISBN 13: 978-1-269-37450-7 British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library Printed in the United States of America ISBN 10: 1-292-02735-5 ISBN 13: 978-1-292-02735-7
- 4. Table of Contents P E A R S O N C U S T O M L I B R A R Y I 1. Optical Fiber Transmission Media 1 Wayne Tomasi 2. Digital Modulation 49 Wayne Tomasi 3. Introduction to Data Communications and Networking 111 Wayne Tomasi 4. Fundamental Concepts of Data Communications 149 Wayne Tomasi 5. Data-Link Protocols and Data Communications Networks 213 Wayne Tomasi 6. Digital Transmission 277 Wayne Tomasi 7. Digital T-Carriers and Multiplexing 323 Wayne Tomasi 8. Telephone Instruments and Signals 383 Wayne Tomasi 9. The Telephone Circuit 405 Wayne Tomasi 10. The Public Telephone Network 439 Wayne Tomasi 11. Cellular Telephone Concepts 469 Wayne Tomasi 12. Cellular Telephone Systems 491 Wayne Tomasi 13. Microwave Radio Communications and System Gain 529 Wayne Tomasi
- 5. II 14. Satellite Communications 565 Wayne Tomasi 609 Index
- 6. CHAPTER OUTLINE 1 Introduction 8 Optical Fiber Configurations 2 History of Optical Fiber Communications 9 Optical Fiber Classifications 3 Optical Fibers versus Metallic Cable Facilities 10 Losses in Optical Fiber Cables 4 Electromagnetic Spectrum 11 Light Sources 5 Block Diagram of an Optical Fiber 12 Optical Sources Communications System 13 Light Detectors 6 Optical Fiber Types 14 Lasers 7 Light Propagation 15 Optical Fiber System Link Budget OBJECTIVES ■ Define optical communications ■ Present an overview of the history of optical fibers and optical fiber communications ■ Compare the advantages and disadvantages of optical fibers over metallic cables ■ Define electromagnetic frequency and wavelength spectrum ■ Describe several types of optical fiber construction ■ Explain the physics of light and the following terms: velocity of propagation, refraction, refractive index, critical angle, acceptance angle, acceptance cone, and numerical aperture ■ Describe how light waves propagate through an optical fiber cable ■ Define modes of propagation and index profile ■ Describe the three types of optical fiber configurations: single-mode step index, multimode step index, and mul- timode graded index ■ Describe the various losses incurred in optical fiber cables ■ Define light source and optical power ■ Describe the following light sources: light-emitting diodes and injection diodes ■ Describe the following light detectors: PIN diodes and avalanche photodiodes ■ Describe the operation of a laser ■ Explain how to calculate a link budget for an optical fiber system Optical Fiber Transmission Media From Chapter 1 of Advanced Electronic Communications Systems, Sixth Edition. Wayne Tomasi. Copyright © 2004 by Pearson Education, Inc. Published by Pearson Prentice Hall. All rights reserved. 1
- 7. 1 INTRODUCTION Optical fiber cables are the newest and probably the most promising type of guided trans- mission medium for virtually all forms of digital and data communications applications, in- cluding local, metropolitan, and wide area networks. With optical fibers, electromagnetic waves are guided through a media composed of a transparent material without using elec- trical current flow. With optical fibers, electromagnetic light waves propagate through the media in much the same way that radio signals propagate through Earth’s atmosphere. In essence, an optical communications system is one that uses light as the carrier of information. Propagating light waves through Earth’s atmosphere is difficult and often im- practical. Consequently, optical fiber communications systems use glass or plastic fiber ca- bles to “contain” the light waves and guide them in a manner similar to the way electro- magnetic waves are guided through a metallic transmission medium. The information-carrying capacity of any electronic communications system is di- rectly proportional to bandwidth. Optical fiber cables have, for all practical purposes, an in- finite bandwidth. Therefore, they have the capacity to carry much more information than their metallic counterparts or, for that matter, even the most sophisticated wireless commu- nications systems. For comparison purposes, it is common to express the bandwidth of an analog com- munications system as a percentage of its carrier frequency. This is sometimes called the bandwidth utilization ratio. For instance, a VHF radio communications system operating at a carrier frequency of 100 MHz with 10-MHz bandwidth has a bandwidth utilization ratio of 10%. A microwave radio system operating at a carrier frequency of 10 GHz with a 10% bandwidth utilization ratio would have 1 GHz of bandwidth available. Obviously, the higher the carrier frequency, the more bandwidth available, and the greater the information- carrying capacity. Light frequencies used in optical fiber communications systems are be- tween 1 1014 Hz and 4 1014 Hz (100,000 GHz to 400,000 GHz). A bandwidth utiliza- tion ratio of 10% would be a bandwidth between 10,000 GHz and 40,000 GHz. 2 HISTORY OF OPTICAL FIBER COMMUNICATIONS In 1880, Alexander Graham Bell experimented with an apparatus he called a photophone. The photophone was a device constructed from mirrors and selenium detectors that trans- mitted sound waves over a beam of light. The photophone was awkward and unreliable and had no real practical application. Actually, visual light was a primary means of communi- cating long before electronic communications came about. Smoke signals and mirrors were used ages ago to convey short, simple messages. Bell’s contraption, however, was the first attempt at using a beam of light for carrying information. Transmission of light waves for any useful distance through Earth’s atmosphere is im- practical because water vapor, oxygen, and particulates in the air absorb and attenuate the signals at light frequencies. Consequently, the only practical type of optical communica- tions system is one that uses a fiber guide. In 1930, J. L. Baird, an English scientist, and C. W. Hansell, a scientist from the United States, were granted patents for scanning and trans- mitting television images through uncoated fiber cables. A few years later, a German sci- entist named H. Lamm successfully transmitted images through a single glass fiber. At that time, most people considered fiber optics more of a toy or a laboratory stunt and, conse- quently, it was not until the early 1950s that any substantial breakthrough was made in the field of fiber optics. In 1951, A. C. S. van Heel of Holland and H. H. Hopkins and N. S. Kapany of En- gland experimented with light transmission through bundles of fibers. Their studies led to the development of the flexible fiberscope, which is used extensively in the medical field. It was Kapany who coined the term “fiber optics” in 1956. Optical Fiber Transmission Media 2
- 8. In 1958, Charles H. Townes, an American, and Arthur L. Schawlow, a Canadian, wrote a paper describing how it was possible to use stimulated emission for amplifying light waves (laser) as well as microwaves (maser). Two years later, Theodore H. Maiman, a sci- entist with Hughes Aircraft Company, built the first optical maser. The laser (light amplification by stimulated emission of radiation) was invented in 1960. The laser’s relatively high output power, high frequency of operation, and capability of carrying an extremely wide bandwidth signal make it ideally suited for high-capacity communications systems. The invention of the laser greatly accelerated research efforts in fiber-optic communications, although it was not until 1967 that K. C. Kao and G. A. Bock- ham of the Standard Telecommunications Laboratory in England proposed a new commu- nications medium using cladded fiber cables. The fiber cables available in the 1960s were extremely lossy (more than 1000 dB/km), which limited optical transmissions to short distances. In 1970, Kapron, Keck, and Maurer of Corning Glass Works in Corning, New York, developed an optical fiber with losses less than 2 dB/km. That was the “big” breakthrough needed to permit practical fiber optics com- munications systems. Since 1970, fiber optics technology has grown exponentially. Re- cently, Bell Laboratories successfully transmitted 1 billion bps through a fiber cable for 600 miles without a regenerator. In the late 1970s and early 1980s, the refinement of optical cables and the development of high-quality, affordable light sources and detectors opened the door to the development of high-quality,high-capacity,efficient,andaffordableopticalfibercommunicationssystems.By the late 1980s, losses in optical fibers were reduced to as low as 0.16 dB/km, and in 1988 NEC Corporation set a new long-haul transmission record by transmitting 10 gigabytes per second over 80.1 kilometers of optical fiber. Also in 1988, the American National Standards Institute (ANSI)publishedtheSynchronousOpticalNetwork(SONET).Bythemid-1990s,opticalvoice and data networks were commonplace throughout the United States and much of the world. 3 OPTICAL FIBERS VERSUS METALLIC CABLE FACILITIES Communications through glass or plastic fibers has several advantages over conven- tional metallic transmission media for both telecommunication and computer networking applications. 3-1 Advantages of Optical Fiber Cables The advantages of using optical fibers include the following: 1. Wider bandwidth and greater information capacity. Optical fibers have greater in- formation capacity than metallic cables because of the inherently wider bandwidths avail- able with optical frequencies. Optical fibers are available with bandwidths up to several thousand gigahertz. The primary electrical constants (resistance, inductance, and capaci- tance) in metallic cables cause them to act like low-pass filters, which limit their transmis- sion frequencies, bandwidth, bit rate, and information-carrying capacity. Modern optical fiber communications systems are capable of transmitting several gigabits per second over hundreds of miles, allowing literally millions of individual voice and data channels to be combined and propagated over one optical fiber cable. 2. Immunity to crosstalk. Optical fiber cables are immune to crosstalk because glass and plastic fibers are nonconductors of electrical current. Therefore, fiber cables are not sur- rounded by a changing magnetic field, which is the primary cause of crosstalk between metallic conductors located physically close to each other. 3. Immunity to static interference. Because optical fiber cables are nonconductors of electrical current, they are immune to static noise due to electromagnetic interference (EMI) caused by lightning, electric motors, relays, fluorescent lights, and other electrical Optical Fiber Transmission Media 3
- 9. noise sources (most of which are man-made). For the same reason, fiber cables do not ra- diate electromagnetic energy. 4. Environmental immunity. Optical fiber cables are more resistant to environmen- tal extremes (including weather variations) than metallic cables. Optical cables also oper- ate over a wider temperature range and are less affected by corrosive liquids and gases. 5. Safety and convenience. Optical fiber cables are safer and easier to install and maintain than metallic cables. Because glass and plastic fibers are nonconductors, there are no electrical currents or voltages associated with them. Optical fibers can be used around volatile liquids and gasses without worrying about their causing explosions or fires. Opti- cal fibers are also smaller and much more lightweight and compact than metallic cables. Consequently, they are more flexible, easier to work with, require less storage space, cheaper to transport, and easier to install and maintain. 6. Lower transmission loss. Optical fibers have considerably less signal loss than their metallic counterparts. Optical fibers are currently being manufactured with as lit- tle as a few-tenths-of-a-decibel loss per kilometer. Consequently, optical regenerators and amplifiers can be spaced considerably farther apart than with metallic transmission lines. 7. Security. Optical fiber cables are more secure than metallic cables. It is virtually impossible to tap into a fiber cable without the user’s knowledge, and optical cables cannot be detected with metal detectors unless they are reinforced with steel for strength. 8. Durability and reliability. Optical fiber cables last longer and are more reliable than metallic facilities because fiber cables have a higher tolerance to changes in environ- mental conditions and are immune to corrosive materials. 9. Economics. The cost of optical fiber cables is approximately the same as metallic cables. Fiber cables have less loss and require fewer repeaters, which equates to lower in- stallation and overall system costs and improved reliability. 3-2 Disadvantages of Optical Fiber Cables Although the advantages of optical fiber cables far exceed the disadvantages, it is impor- tant to know the limitations of the fiber. The disadvantages of optical fibers include the following: 1. Interfacing costs. Optical fiber cable systems are virtually useless by themselves. To be practical and useful, they must be connected to standard electronic facilities, which often require expensive interfaces. 2. Strength. Optical fibers by themselves have a significantly lower tensile strength than coaxial cable. This can be improved by coating the fiber with standard Kevlar and a protective jacket of PVC. In addition, glass fiber is much more fragile than copper wire, making fiber less attractive where hardware portability is required. 3. Remote electrical power. Occasionally, it is necessary to provide electrical power to remote interface or regenerating equipment. This cannot be accomplished with the opti- cal cable, so additional metallic cables must be included in the cable assembly. 4. Optical fiber cables are more susceptible to losses introduced by bending the ca- ble. Electromagnetic waves propagate through an optical cable by either refraction or re- flection. Therefore, bending the cable causes irregularities in the cable dimensions, result- ing in a loss of signal power. Optical fibers are also more prone to manufacturing defects, as even the most minor defect can cause excessive loss of signal power. 5. Specialized tools, equipment, and training. Optical fiber cables require special tools to splice and repair cables and special test equipment to make routine measurements. Not only is repairing fiber cables difficult and expensive, but technicians working on opti- cal cables also require special skills and training. In addition, sometimes it is difficult to lo- cate faults in optical cables because there is no electrical continuity. Optical Fiber Transmission Media 4
- 10. 100 Hz Subsonic Audio AM radio FM radio and television Terrestrial microwave satelite and radar Infrared light Visible light Ultraviolet light X-rays Gamma rays Cosmic rays Ultrasonic 101 102 103 104 105 106 107 108 109 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 kHz (kilo) MHz (mega) GHz (giga) THz (tera) Frequency PHz (penta) EHz (exa) FIGURE 1 Electromagnetic frequency spectrum 4 ELECTROMAGNETIC SPECTRUM The total electromagnetic frequency spectrum is shown in Figure 1. From the figure, it can be seen that the frequency spectrum extends from the subsonic frequencies (a few hertz) to cos- mic rays (1022 Hz). The light frequency spectrum can be divided into three general bands: 1. Infrared. The band of light frequencies that is too high to be seen by the human eye with wavelengths ranging between 770 nm and 106 nm. Optical fiber systems generally operate in the infrared band. 2. Visible.Thebandoflightfrequenciestowhichthehumaneyewillrespondwithwave- lengths ranging between 390 nm and 770 nm. This band is visible to the human eye. 3. Ultraviolet. The band of light frequencies that are too low to be seen by the hu- man eye with wavelengths ranging between 10 nm and 390 nm. When dealing with ultra-high-frequency electromagnetic waves, such as light, it is common to use units of wavelength rather than frequency. Wavelength is the length that one cycle of an electromagnetic wave occupies in space. The length of a wavelength depends on the frequency of the wave and the velocity of light. Mathematically, wavelength is (1) where λ wavelength (meters/cycle) c velocity of light (300,000,000 meters per second) f frequency (hertz) With light frequencies, wavelength is often stated in microns, where 1 micron 106 meter (1 μm), or in nanometers (nm), where 1 nm 109 meter. However, when describ- ing the optical spectrum, the unit angstrom is sometimes used to express wavelength, where 1 angstrom 1010 meter, or 0.0001 micron. Figure 2 shows the total electromagnetic wavelength spectrum. 5 BLOCK DIAGRAM OF AN OPTICAL FIBER COMMUNICATIONS SYSTEM Figure 3 shows a simplified block diagram of a simplex optical fiber communications link. The three primary building blocks are the transmitter, the receiver, and the optical fiber ca- ble. The transmitter is comprised of a voltage-to-current converter, a light source, and a source-to-fiber interface (light coupler). The fiber guide is the transmission medium, which λ c f Optical Fiber Transmission Media 5
- 11. 10-7 Hz 10-6 10-5 10-4 10-3 10-2 10-1 100 101 102 103 104 105 106 107 108 109 1010 1011 1012 1013 1014 1015 1016 1017 Wavelength Gamma rays μm 0.01 Å 100 nm 10 2 2000 200 3 3000 300 3.9 3900 390 4.55 4550 455 4.92 4920 492 5.77 5770 577 5.97 5970 597 6.22 6220 622 7.7 7700 770 15 15,000 1500 60 60,000 6000 400 400,000 40000 1000 1,000,000 10000 Long electrical oscillations X-rays Extreme Far Ultraviolet Visible light Infrared Near Vio Blue Green Yel Orng Red Near Middle Far Far Far Cosmic rays Microwaves Radio waves FIGURE 2 Electromagnetic wavelength spectrum Source-to- fiber interface Voltage-to- current converter Light source Analog or digital interface Destination Fiber-to- light detector interface Light detector Signal regenerator Analog or digital interface Source Transmitter Optical fiber cable Optical fiber cable Receiver Current-to- voltage converter FIGURE 3 Optical fiber communications link 6
- 12. Polyurethane outer jacket Strength members Buffer jacket Fiber core and cladding Protective coating FIGURE 4 Optical fiber cable construction is either an ultrapure glass or a plastic cable. It may be necessary to add one or more re- generators to the transmission medium, depending on the distance between the transmitter and receiver. Functionally, the regenerator performs light amplification. However, in real- ity the signal is not actually amplified; it is reconstructed. The receiver includes a fiber-to- interface (light coupler), a photo detector, and a current-to-voltage converter. In the transmitter, the light source can be modulated by a digital or an analog signal. The voltage-to-current converter serves as an electrical interface between the input circuitry and the light source. The light source is either an infrared light-emitting diode (LED) or an injection laser diode (ILD). The amount of light emitted by either an LED or ILD is pro- portional to the amount of drive current. Thus, the voltage-to-current converter converts an input signal voltage to a current that is used to drive the light source. The light outputted by the light source is directly proportional to the magnitude of the input voltage. In essence, the light intensity is modulated by the input signal. The source-to-fiber coupler (such as an optical lens) is a mechanical interface. Its function is to couple light emitted by the light source into the optical fiber cable. The opti- cal fiber consists of a glass or plastic fiber core surrounded by a cladding and then encap- sulated in a protective jacket. The fiber-to-light detector-coupling device is also a mechan- ical coupler. Its function is to couple as much light as possible from the fiber cable into the light detector. The light detector is generally a PIN (p-type-intrinsic-n-type) diode, an APD (ava- lanche photodiode), or a phototransistor. All three of these devices convert light energy to current. Consequently, a current-to-voltage converter is required to produce an output volt- age proportional to the original source information. The current-to-voltage converter trans- forms changes in detector current to changes in voltage. The analog or digital interfaces are electrical interfaces that match impedances and signal levels between the information source and destination to the input and output cir- cuitry of the optical system. 6 OPTICAL FIBER TYPES 6-1 Optical Fiber Construction The actual fiber portion of an optical cable is generally considered to include both the fiber core and its cladding (see Figure 4). A special lacquer, silicone, or acrylate coating is gen- erally applied to the outside of the cladding to seal and preserve the fiber’s strength, help- Optical Fiber Transmission Media 7
- 13. ing maintain the cables attenuation characteristics. The coating also helps protect the fiber from moisture, which reduces the possibility of the occurrence of a detrimental phenome- non called stress corrosion (sometimes called static fatigue) caused by high humidity. Moisture causes silicon dioxide crystals to interact, causing bonds to break down and spon- taneous fractures to form over a prolonged period of time. The protective coating is sur- rounded by a buffer jacket, which provides the cable additional protection against abrasion and shock. Materials commonly used for the buffer jacket include steel, fiberglass, plastic, flame-retardant polyvinyl chloride (FR-PVC), Kevlar yarn, and paper. The buffer jacket is encapsulated in a strength member, which increases the tensile strength of the overall cable assembly. Finally, the entire cable assembly is contained in an outer polyurethane jacket. There are three essential types of optical fibers commonly used today. All three vari- eties are constructed of either glass, plastic, or a combination of glass and plastic: Plastic core and cladding Glass core with plastic cladding (called PCS fiber [plastic-clad silica]) Glass core and glass cladding (called SCS [silica-clad silica]) Plastic fibers are more flexible and, consequently, more rugged than glass. Therefore, plastic cables are easier to install, can better withstand stress, are less expensive, and weigh approximately 60% less than glass. However, plastic fibers have higher attenuation charac- teristics and do not propagate light as efficiently as glass. Therefore, plastic fibers are lim- ited to relatively short cable runs, such as within a single building. Fibers with glass cores have less attenuation than plastic fibers, with PCS being slightly better than SCS. PCS fibers are also less affected by radiation and, therefore, are more immune to external interference. SCS fibers have the best propagation characteristics and are easier to terminate than PCS fibers. Unfortunately, SCS fibers are the least rugged, and they are more susceptible to increases in attenuation when exposed to radiation. The selection of a fiber for a given application is a function of the specific system re- quirements. There are always trade-offs based on the economics and logistics of a particu- lar application. 6-1-1 Cable configurations. There are many different cable designs available today. Figure 5 shows examples of several optical fiber cable configurations. With loose tube con- struction (Figure 5a), each fiber is contained in a protective tube. Inside the tube, a polyurethane compound encapsules the fiber and prevents the intrusion of water. A phe- nomenon called stress corrosion or static fatigue can result if the glass fiber is exposed to long periods of high humidity. Silicon dioxide crystals interact with the moisture and cause bonds to break down, causing spontaneous fractures to form over a prolonged period. Some fiber cables have more than one protective coating to ensure that the fiber’s characteristics do not alter if the fiber is exposed to extreme temperature changes. Surrounding the fiber’s cladding is usually a coating of either lacquer, silicon, or acrylate that is typically applied to seal and preserve the fiber’s strength and attenuation characteristics. Figure 5b shows the construction of a constrained optical fiber cable. Surrounding the fiber are a primary and a secondary buffer comprised of Kevlar yarn, which increases the tensile strength of the cable and provides protection from external mechanical influences that could cause fiber breakage or excessive optical attenuation. Again, an outer protective tube is filled with polyurethane, which prevents moisture from coming into contact with the fiber core. Figure 5c shows a multiple-strand cable configuration, which includes a steel central member and a layer of Mylar tape wrap to increase the cable’s tensile strength. Figure 5d shows a ribbon configuration for a telephone cable, and Figure 5e shows both end and side views of a PCS cable. Optical Fiber Transmission Media 8
- 14. FIGURE 5 Fiber optic cable configurations: (a) loose tube construction; (b) constrained fiber; (c) multiple strands; (d) telephone cable; (e) plastic-silica cable As mentioned, one disadvantage of optical fiber cables is their lack of tensile (pulling) strength, which can be as low as a pound. For this reason, the fiber must be rein- forced with strengthening material so that it can withstand mechanical stresses it will typi- cally undergo when being pulled and jerked through underground and overhead ducts and hung on poles. Materials commonly used to strengthen and protect fibers from abrasion and environmental stress are steel, fiberglass, plastic, FR-PVC (flame-retardant polyvinyl chlo- ride), Kevlar yarn, and paper. The type of cable construction used depends on the perfor- mance requirements of the system and both economic and environmental constraints. 7 LIGHT PROPAGATION 7-1 The Physics of Light Although the performance of optical fibers can be analyzed completely by application of Maxwell’s equations, this is necessarily complex. For most practical applications, geomet- ric wave tracing may be used instead. Optical Fiber Transmission Media 9
- 15. In 1860, James Clerk Maxwell theorized that electromagnetic radiation contained a series of oscillating waves comprised of an electric and a magnetic field in quadrature (at 90° angles). However, in 1905, Albert Einstein and Max Planck showed that when light is emitted or absorbed, it behaves like an electromagnetic wave and also like a particle, called a photon, which possesses energy proportional to its frequency. This theory is known as Planck’s law. Planck’s law describes the photoelectric effect, which states, “When visible light or high-frequency electromagnetic radiation illuminates a metallic surface, electrons are emitted.” The emitted electrons produce an electric current. Planck’s law is expressed mathematically as Ep hf (2) where Ep energy of the photon (joules) h Planck’s constant 6.625 1034 J s f frequency of light (photon) emitted (hertz) Photon energy may also be expressed in terms of wavelength. Substituting Equation 1 into Equation 2 yields Ep hf (3a) or (3b) An atom has several energy levels or states, the lowest of which is the ground state. Any energy level above the ground state is called an excited state. If an atom in one energy level decays to a lower energy level, the loss of energy (in electron volts) is emitted as a photon of light. The energy of the photon is equal to the difference between the energy of the two energy levels. The process of decaying from one energy level to another energy level is called spontaneous decay or spontaneous emission. Atoms can be irradiated by a light source whose energy is equal to the difference be- tween ground level and an energy level. This can cause an electron to change from one en- ergy level to another by absorbing light energy. The process of moving from one energy level to another is called absorption. When making the transition from one energy level to another, the atom absorbs a packet of energy (a photon). This process is similar to that of emission. The energy absorbed or emitted (photon) is equal to the difference between the two energy levels. Mathematically, Ep E2 E1 (4) where Ep is the energy of the photon (joules). 7-2 Optical Power Light intensity is a rather complex concept that can be expressed in either photometric or radiometric terms. Photometry is the science of measuring only light waves that are visible to the human eye. Radiometry, on the other hand, measures light throughout the entire elec- tromagnetic spectrum. In photometric terms, light intensity is generally described in terms of luminous flux density and measured in lumens per unit area. Radiometric terms, how- ever, are often more useful to engineers and technologists. In radiometric terms, optical power measures the rate at which electromagnetic waves transfer light energy. In simple terms, optical power is described as the flow of light energy past a given point in a speci- fied time. Optical power is expressed mathematically as (5a) P d1energy2 d1time2 Ep hc λ Optical Fiber Transmission Media 10
- 16. or (5b) where P optical power (watts) dQ instantaneous charge (joules) dt instantaneous change in time (seconds) Optical power is sometimes called radiant flux (φ), which is equivalent to joules per second and is the same power that is measured electrically or thermally in watts. Radio- metric terms are generally used with light sources with output powers ranging from tens of microwatts to more than 100 milliwatts. Optical power is generally stated in decibels relative to a defined power level, such as 1 mW (dBm) or 1 μW (dBμ). Mathematically stated, (6) and (7) Example 1 Determine the optical power in dBm and dBμ for power levels of a. 10 mW b. 20 μW Solution a. Substituting into Equations 6 and 7 gives b. Substituting into Equations 6 and 7 gives 7-3 Velocity of Propagation In free space (a vacuum), electromagnetic energy, such as light waves, travels at ap- proximately 300,000,000 meters per second (186,000 mi/s). Also, in free space the ve- locity of propagation is the same for all light frequencies. However, it has been demon- strated that electromagnetic waves travel slower in materials more dense than free space and that all light frequencies do not propagate at the same velocity. When the ve- locity of an electromagnetic wave is reduced as it passes from one medium to another medium of denser material, the light ray changes direction or refracts (bends) toward the normal. When an electromagnetic wave passes from a more dense material into a less dense material, the light ray is refracted away from the normal. The normal is sim- ply an imaginary line drawn perpendicular to the interface of the two materials at the point of incidence. dBμ 10 log 20 μW 1μW 13 dBμ dBm 10 log 20 μW 1 mW 17 dBm dBμ 10 log 10 mW 1 μW 40 dBμ dBm 10 log 10 mW 1 mW 10 dBm dbμ 10 logB P 1watts2 0.000001 1watts2 R dBm 10 logB P 1watts2 0.001 1watts2 R dQ dt Optical Fiber Transmission Media 11
- 17. FIGURE 6 Refraction of light: (a) light refraction; (b) prismatic refraction 7-3-1 Refraction. For light-wave frequencies, electromagnetic waves travel through Earth’s atmosphere (air) at approximately the same velocity as through a vacuum (i.e., the speed of light). Figure 6a shows how a light ray is refracted (bent) as it passes from a less dense material into a more dense material. (Actually, the light ray is not bent; rather, it changes direction at the interface.) Figure 6b shows how sunlight, which contains all light frequencies (white light), is affected as it passes through a material that is more dense than air. Refraction occurs at both air/glass interfaces. The violet wavelengths are refracted the most, whereas the red wavelengths are refracted the least. The spectral separation of white light in this manner is called prismatic refraction. It is this phenomenon that causes rain- bows, where water droplets in the atmosphere act as small prisms that split the white sun- light into the various wavelengths, creating a visible spectrum of color. 7-3-2 Refractive Index. The amount of bending or refraction that occurs at the in- terface of two materials of different densities is quite predictable and depends on the re- fractive indexes of the two materials. Refractive index is simply the ratio of the velocity of propagation of a light ray in free space to the velocity of propagation of a light ray in a given material. Mathematically, refractive index is (8) n c v Optical Fiber Transmission Media 12
- 18. Table 1 Typical Indexes of Refraction Material Index of Refractiona Vacuum 1.0 Air 1.0003 (≈1) Water 1.33 Ethyl alcohol 1.36 Fused quartz 1.46 Glass fiber 1.5–1.9 Diamond 2.0–2.42 Silicon 3.4 Gallium-arsenide 2.6 a Index of refraction is based on a wavelength of light emitted from a sodium flame (589 nm). FIGURE 7 Refractive model for Snell’s law where n refractive index (unitless) c speed of light in free space (3 108 meters per second) v speed of light in a given material (meters per second) Although the refractive index is also a function of frequency, the variation in most light wave applications is insignificant and, thus, omitted from this discussion. The indexes of refraction of several common materials are given in Table 1. 7-3-3 Snell’s law. How a light ray reacts when it meets the interface of two trans- missive materials that have different indexes of refraction can be explained with Snell’s law. A refractive index model for Snell’s law is shown in Figure 7. The angle of incidence is the angle at which the propagating ray strikes the interface with respect to the normal, and the angle of refraction is the angle formed between the propagating ray and the nor- mal after the ray has entered the second medium.At the interface of medium 1 and medium 2, the incident ray may be refracted toward the normal or away from it, depending on whether n1 is greater than or less than n2. Hence, the angle of refraction can be larger or Optical Fiber Transmission Media 13
- 19. FIGURE 8 Light ray refracted away from the normal smaller than the angle of incidence, depending on the refractive indexes of the two materi- als. Snell’s law stated mathematically is n1 sin θ1 n2 sin θ2 (9) where n1 refractive index of material 1 (unitless) n2 refractive index of material 2 (unitless) θ1 angle of incidence (degrees) θ2 angle of refraction (degrees) Figure 8 shows how a light ray is refracted as it travels from a more dense (higher refractive index) material into a less dense (lower refractive index) material. It can be seen that the light ray changes direction at the interface, and the angle of refraction is greater than the angle of incidence. Consequently, when a light ray enters a less dense material, the ray bends away from the normal. The normal is simply a line drawn per- pendicular to the interface at the point where the incident ray strikes the interface. Similarly, when a light ray enters a more dense material, the ray bends toward the normal. Example 2 In Figure 8, let medium 1 be glass and medium 2 be ethyl alcohol. For an angle of incidence of 30°, determine the angle of refraction. Solution From Table 1, n1 (glass) 1.5 n2 (ethyl alcohol) 1.36 Rearranging Equation 9 and substituting for n1, n2, and θ1 gives us The result indicates that the light ray refracted (bent) or changed direction by 33.47° at the interface. Because the light was traveling from a more dense material into a less dense material, the ray bent away from the normal. θ2 sin1 0.5514 33.47° 1.5 1.36 sin 30 0.5514 sin θ2 n1 n2 sin θ1 sin θ2 Optical Fiber Transmission Media 14
- 20. FIGURE 9 Critical angle refraction 7-3-4 Critical angle. Figure 9 shows a condition in which an incident ray is strik- ing the glass/cladding interface at an angle (1) such that the angle of refraction (θ2) is 90° and the refracted ray is along the interface. This angle of incidence is called the critical an- gle (θc), which is defined as the minimum angle of incidence at which a light ray may strike the interface of two media and result in an angle of refraction of 90° or greater. It is impor- tant to note that the light ray must be traveling from a medium of higher refractive index to a medium with a lower refractive index (i.e., glass into cladding). If the angle of refraction is 90° or greater, the light ray is not allowed to penetrate the less dense material. Conse- quently, total reflection takes place at the interface, and the angle of reflection is equal to the angle of incidence. Critical angle can be represented mathematically by rearranging Equation 9 as With θ2 90°, θ1 becomes the critical angle (θc), and and (10) where θc is the critical angle. From Equation 10, it can be seen that the critical angle is dependent on the ratio of the refractive indexes of the core and cladding. For example a ratio n2 /n1 0.77 produces a critical angle of 50.4°, whereas a ratio n2 /n1 0.625 yields a critical angle of 38.7°. Figure 10 shows a comparison of the angle of refraction and the angle of reflection when the angle of incidence is less than or more than the critical angle. 7-3-5 Acceptance angle, acceptance cone, and numerical aperture. In previous discussions, the source-to-fiber aperture was mentioned several times, and the critical and acceptance angles at the point where a light ray strikes the core/cladding interface were ex- plained. The following discussion addresses the light-gathering ability of a fiber, which is the ability to couple light from the source into the fiber. θc sin1 n2 n1 sin θc n2 n1 112 sin θc n2 n1 sin θ1 n2 n1 sin θ2 Optical Fiber Transmission Media 15
- 21. FIGURE 10 Angle of reflection and refraction FIGURE 11 Ray propagation into and down an optical fiber cable Figure 11 shows the source end of a fiber cable and a light ray propagating into and then down the fiber. When light rays enter the core of the fiber, they strike the air/glass in- terface at normal A. The refractive index of air is approximately 1, and the refractive index of the glass core is 1.5. Consequently, the light enters the cable traveling from a less dense to a more dense medium, causing the ray to refract toward the normal. This causes the light rays to change direction and propagate diagonally down the core at an angle that is less than the external angle of incidence (θin). For a ray of light to propagate down the cable, it must strike the internal core/cladding interface at an angle that is greater than the critical angle (θc). Using Figure 12 and Snell’s law, it can be shown that the maximum angle that exter- nal light rays may strike the air/glass interface and still enter the core and propagate down the fiber is Optical Fiber Transmission Media 16
- 22. FIGURE 12 Geometric relationship of Equations 11a and b (11a) where θin(max) acceptance angle (degrees) no refractive index of air (1) n1 refractive index of glass fiber core (1.5) n2 refractive index of quartz fiber cladding (1.46) Since the refractive index of air is 1, Equation 11a reduces to (11b) θin(max) is called the acceptance angle or acceptance cone half-angle. θin(max) defines the maximum angle in which external light rays may strike the air/glass interface and still propagate down the fiber. Rotating the acceptance angle around the fiber core axis de- scribes the acceptance cone of the fiber input. Acceptance cone is shown in Figure 13a, and the relationship between acceptance angle and critical angle is shown in Figure 13b. Note that the critical angle is defined as a minimum value and that the acceptance angle is de- fined as a maximum value. Light rays striking the air/glass interface at an angle greater than the acceptance angle will enter the cladding and, therefore, will not propagate down the cable. Numerical aperture (NA) is closely related to acceptance angle and is the figure of merit commonly used to measure the magnitude of the acceptance angle. In essence, nu- merical aperture is used to describe the light-gathering or light-collecting ability of an op- tical fiber (i.e., the ability to couple light into the cable from an external source). The larger the magnitude of the numerical aperture, the greater the amount of external light the fiber will accept. The numerical aperture for light entering the glass fiber from an air medium is described mathematically as NA sin θin (12a) and (12b) Therefore θin sin1 NA (12c) where θin acceptance angle (degrees) NA numerical aperture (unitless) n1 refractive index of glass fiber core (unitless) n2 refractive index of quartz fiber cladding (unitless) NA 2n1 2 n2 2 θin1max2 sin1 2n1 2 n2 2 θin1max2 sin1 2n1 2 n2 2 no Optical Fiber Transmission Media 17
- 23. FIGURE 13 (a) Acceptance angle; (b) acceptance cone A larger-diameter core does not necessarily produce a larger numerical aperture, al- though in practice larger-core fibers tend to have larger numerical apertures. Numerical aperture can be calculated using Equations 12a or b, but in practice it is generally measured by looking at the output of a fiber because the light-guiding properties of a fiber cable are symmetrical. Therefore, light leaves a cable and spreads out over an angle equal to the ac- ceptance angle. 8 OPTICAL FIBER CONFIGURATIONS Light can be propagated down an optical fiber cable using either reflection or refraction. How the light propagates depends on the mode of propagation and the index profile of the fiber. 8-1 Mode of Propagation In fiber optics terminology, the word mode simply means path. If there is only one path for light rays to take down a cable, it is called single mode. If there is more than one path, it is called multimode. Figure 14 shows single and multimode propagation of light rays down an optical fiber. As shown in Figure 14a, with single-mode propagation, there is only one Optical Fiber Transmission Media 18
- 24. FIGURE 14 Modes of propagation: (a) single mode; (b) multimode path for light rays to take, which is directly down the center of the cable. However, as Figure 14b shows, with multimode propagation there are many higher-order modes possible, and light rays propagate down the cable in a zigzag fashion following several paths. The number of paths (modes) possible for a multimode fiber cable depends on the fre- quency (wavelength) of the light signal, the refractive indexes of the core and cladding, and the core diameter. Mathematically, the number of modes possible for a given cable can be approximated by the following formula: (13) where N number of propagating modes d core diameter (meters) λ wavelength (meters) n1 refractive index of core n2 refractive index of cladding A multimode step-index fiber with a core diameter of 50 μm, a core refractive index of 1.6, a cladding refractive index of 1.584, and a wavelength of 1300 nm has approximately 372 possible modes. 8-2 Index Profile The index profile of an optical fiber is a graphical representation of the magnitude of the refractive index across the fiber. The refractive index is plotted on the horizontal axis, and the radial distance from the core axis is plotted on the vertical axis. Figure 15 shows the core index profiles for the three types of optical fiber cables. There are two basic types of index profiles: step and graded. A step-index fiber has a central core with a uniform refractive index (i.e., constant density throughout). An outside cladding that also has a uniform refractive index surrounds the core; however, the refractive index of the cladding is less than that of the central core. From Figures 15a and b, it can be seen that in step-index fibers, there is an abrupt change in the refractive index at the core/cladding interface. This is true for both single and multimode step-index fibers. N ⬇ ¢ πd λ 2n1 2 n2 2 ≤ 2 Optical Fiber Transmission Media 19
- 25. FIGURE 15 Core index profiles: (a) single-mode step index; (b) multimode step index; (c) multimode graded index In the graded-index fiber, shown in Figure 15c, it can be see that there is no cladding, and the refractive index of the core is nonuniform; it is highest in the center of the core and decreases gradually with distance toward the outer edge. The index profile shows a core density that is maximum in the center and decreases symmetrically with distance from the center. 9 OPTICAL FIBER CLASSIFICATIONS Propagation modes can be categorized as either multimode or single mode, and then mul- timode can be further subdivided into step index or graded index. Although there are a wide variety of combinations of modes and indexes, there are only three practical types of opti- cal fiber configurations: single-mode step-index, multimode step index, and multimode graded index. 9-1 Single-Mode Step-Index Optical Fiber Single-mode step-index fibers are the dominant fibers used in today’s telecommunications and data networking industries.A single-mode step-index fiber has a central core that is sig- nificantly smaller in diameter than any of the multimode cables. In fact, the diameter is suf- ficiently small that there is essentially only one path that light may take as it propagates down the cable. This type of fiber is shown in Figure 16a. In the simplest form of single- mode step-index fiber, the outside cladding is simply air. The refractive index of the glass core (n1) is approximately 1.5, and the refractive index of the air cladding (n2) is 1. The large Optical Fiber Transmission Media 20
- 26. FIGURE 16 Single-mode step-index fibers: (a) air cladding; (b) glass cladding difference in the refractive indexes results in a small critical angle (approximately 42°) at the glass/air interface. Consequently, a single-mode step-index fiber has a wide external ac- ceptance angle, which makes it relatively easy to couple light into the cable from an exter- nal source. However, this type of fiber is very weak and difficult to splice or terminate. A more practical type of single-mode step-index fiber is one that has a cladding other than air, such as the cable shown in Figure 16b. The refractive index of the cladding (n2) is slightly less than that of the central core (n1) and is uniform throughout the cladding. This type of cable is physically stronger than the air-clad fiber, but the critical angle is also much higher (approximately 77°). This results in a small acceptance angle and a narrow source-to- fiber aperture, making it much more difficult to couple light into the fiber from a light source. With both types of single-mode step-index fibers, light is propagated down the fiber through reflection. Light rays that enter the fiber either propagate straight down the core or, perhaps, are reflected only a few times. Consequently, all light rays follow approximately the same path down the cable and take approximately the same amount of time to travel the length of the cable. This is one overwhelming advantage of single-mode step-index fibers, as explained in more detail in a later section of this chapter. 9-2 Multimode Step-Index Optical Fiber A multimode step-index optical fiber is shown in Figure 17. Multimode step-index fibers are similar to the single-mode step-index fibers except the center core is much larger with the multimode configuration. This type of fiber has a large light-to-fiber aperture and, con- sequently, allows more external light to enter the cable. The light rays that strike the core/cladding interface at an angle greater than the critical angle (ray A) are propagated down the core in a zigzag fashion, continuously reflecting off the interface boundary. Light Optical Fiber Transmission Media 21
- 27. FIGURE 18 Multimode graded-index fiber FIGURE 17 Multimode step-index fiber rays that strike the core/cladding interface at an angle less than the critical angle (ray B) en- ter the cladding and are lost. It can be seen that there are many paths that a light ray may follow as it propagates down the fiber. As a result, all light rays do not follow the same path and, consequently, do not take the same amount of time to travel the length of the cable. 9-3 Multimode Graded-Index Optical Fiber A multimode graded-index optical fiber is shown in Figure 18. Graded-index fibers are char- acterized by a central core with a nonuniform refractive index. Thus, the cable’s density is maximum at the center and decreases gradually toward the outer edge. Light rays propagate down this type of fiber through refraction rather than reflection.As a light ray propagates di- agonally across the core toward the center, it is continually intersecting a less dense to more dense interface. Consequently, the light rays are constantly being refracted, which results in a continuous bending of the light rays. Light enters the fiber at many different angles.As the light rays propagate down the fiber, the rays traveling in the outermost area of the fiber travel a greater distance than the rays traveling near the center. Because the refractive index de- creases with distance from the center and the velocity is inversely proportional to refractive index, the light rays traveling farthest from the center propagate at a higher velocity. Conse- quently, they take approximately the same amount of time to travel the length of the fiber. 9-4 Optical Fiber Comparison 9-4-1 Single-mode step-index fiber. Advantages include the following: 1. Minimum dispersion: All rays propagating down the fiber take approximately the same path; thus, they take approximately the same length of time to travel down the cable. Consequently, a pulse of light entering the cable can be reproduced at the receiving end very accurately. Optical Fiber Transmission Media 22
- 28. 2. Because of the high accuracy in reproducing transmitted pulses at the receive end, wider bandwidths and higher information transmission rates (bps) are possible with single-mode step-index fibers than with the other types of fibers. Disadvantages include the following: 1. Because the central core is very small, it is difficult to couple light into and out of this type of fiber. The source-to-fiber aperture is the smallest of all the fiber types. 2. Again, because of the small central core, a highly directive light source, such as a laser, is required to couple light into a single-mode step-index fiber. 3. Single-mode step-index fibers are expensive and difficult to manufacture. 9-4-2 Multimode step-index fiber. Advantages include the following: 1. Multimode step-index fibers are relatively inexpensive and simple to manufacture. 2. It is easier to couple light into and out of multimode step-index fibers because they have a relatively large source-to-fiber aperture. Disadvantages include the following: 1. Light rays take many different paths down the fiber, which results in large dif- ferences in propagation times. Because of this, rays traveling down this type of fiber have a tendency to spread out. Consequently, a pulse of light propagating down a multimode step-index fiber is distorted more than with the other types of fibers. 2. The bandwidths and rate of information transfer rates possible with this type of cable are less than that possible with the other types of fiber cables. 9-4-3 Multimode graded-index fiber. Essentially, there are no outstanding advan- tages or disadvantages of this type of fiber. Multimode graded-index fibers are easier to cou- ple light into and out of than single-mode step-index fibers but are more difficult than mul- timode step-index fibers. Distortion due to multiple propagation paths is greater than in single-mode step-index fibers but less than in multimode step-index fibers. This multimode graded-index fiber is considered an intermediate fiber compared to the other fiber types. 10 LOSSES IN OPTICAL FIBER CABLES Power loss in an optical fiber cable is probably the most important characteristic of the ca- ble. Power loss is often called attenuation and results in a reduction in the power of the light wave as it travels down the cable. Attenuation has several adverse effects on performance, including reducing the system’s bandwidth, information transmission rate, efficiency, and overall system capacity. The standard formula for expressing the total power loss in an optical fiber cable is (14) where A(dB) total reduction in power level, attenuation (unitless) Pout cable output power (watts) Pin cable input power (watts) In general, multimode fibers tend to have more attenuation than single-mode cables, primarily because of the increased scattering of the light wave produced from the dopants in the glass. Table 2 shows output power as a percentage of input power for an optical A1dB2 10 log¢ Pout Pin ≤ Optical Fiber Transmission Media 23
- 29. Table 3 Fiber Cable Attenuation Core Diameter Cladding Diameter NA Attenuation Cable Type (μm) (μm) (unitless) (dB/km) Single mode 8 125 — 0.5 at 1300 nm 5 125 — 0.4 at 1300 nm Graded index 50 125 0.2 4 at 850 nm 100 140 0.3 5 at 850 nm Step index 200 380 0.27 6 at 850 nm 300 440 0.27 6 at 850 nm PCS 200 350 0.3 10 at 790 nm 400 550 0.3 10 at 790 nm Plastic — 750 0.5 400 at 650 nm — 1000 0.5 400 at 650 nm Table 2 % Output Power versus Loss in dB Loss (dB) Output Power (%) 1 79 3 50 6 25 9 12.5 10 10 13 5 20 1 30 0.1 40 0.01 50 0.001 fiber cable with several values of decibel loss. A 1-dB cable loss reduces the output power to 50% of the input power. Attenuation of light propagating through glass depends on wavelength. The three wavelength bands typically used for optical fiber communications systems are centered around 0.85 microns, 1.30 microns, and 1.55 microns. For the kind of glass typically used for optical communications systems, the 1.30-micron and 1.55-micron bands have less than 5% loss per kilometer, while the 0.85-micron band experiences almost 20% loss per kilometer. Although total power loss is of primary importance in an optical fiber cable, attenu- ation is generally expressed in decibels of loss per unit length. Attenuation is expressed as a positive dB value because by definition it is a loss. Table 3 lists attenuation in dB/km for several types of optical fiber cables. The optical power in watts measured at a given distance from a power source can be determined mathematically as P Pt 10Al/10 (15) where P measured power level (watts) Pt transmitted power level (watts) A cable power loss (dB/km) l cable length (km) Likewise, the optical power in decibel units is P(dBm) Pin(dBm) Al(dB) (16) where P measured power level (dBm) Pin transmit power (dBm) Al cable power loss, attenuation (dB) Optical Fiber Transmission Media 24
- 30. Example 3 For a single-mode optical cable with 0.25-dB/km loss, determine the optical power 100 km from a 0.1-mW light source. Solution Substituting into Equation 15 gives P 0.1mW 10{[(0.25)(100)]/(10)} 1 104 10{[(0.25)(100)]/(10)} (1 104 )(1 102.5 ) 0.316 μW and or by substituting into Equation 16 Transmission losses in optical fiber cables are one of the most important characteristics of the fibers. Losses in the fiber result in a reduction in the light power, thus reducing the sys- tem bandwidth, information transmission rate, efficiency, and overall system capacity. The predominant losses in optical fiber cables are the following: Absorption loss Material, or Rayleigh, scattering losses Chromatic, or wavelength, dispersion Radiation losses Modal dispersion Coupling losses 10-1 Absorption Losses Absorption losses in optical fibers is analogous to power dissipation in copper cables; im- purities in the fiber absorb the light and convert it to heat. The ultrapure glass used to man- ufacture optical fibers is approximately 99.9999% pure. Still, absorption losses between 1 dB/km and 1000 dB/km are typical. Essentially, there are three factors that contribute to the absorption losses in optical fibers: ultraviolet absorption, infrared absorption, and ion res- onance absorption. 10-1-1 Ultraviolet absorption. Ultraviolet absorption is caused by valence elec- trons in the silica material from which fibers are manufactured. Light ionizes the valence electrons into conduction. The ionization is equivalent to a loss in the total light field and, consequently, contributes to the transmission losses of the fiber. 10-1-2 Infrared absorption. Infrared absorption is a result of photons of light that are absorbed by the atoms of the glass core molecules. The absorbed photons are converted to random mechanical vibrations typical of heating. 10-1-3 Ion resonance absorption. Ion resonance absorption is caused by OH ions in the material. The source of the OH ions is water molecules that have been trapped in the glass during the manufacturing process. Iron, copper, and chromium molecules also cause ion absorption. 35 dBm 10 dBm 25 dB P1dBm2 10 log¢ 0.1 mW 0.001 W ≤ 31100 km210.25 dBkm2 4 35 dBm P1dBm2 10 log¢ 0.316 μW 0.001 ≤ Optical Fiber Transmission Media 25
- 31. FIGURE 19 Absorption losses in optical fibers Figure 19 shows typical losses in optical fiber cables due to ultraviolet, infrared, and ion resonance absorption. 10-2 Material, or Rayleigh, Scattering Losses During manufacturing, glass is drawn into long fibers of very small diameter. During this process, the glass is in a plastic state (not liquid and not solid). The tension applied to the glass causes the cooling glass to develop permanent submicroscopic irregularities. When light rays propagating down a fiber strike one of these impurities, they are diffracted. Dif- fraction causes the light to disperse or spread out in many directions. Some of the dif- fracted light continues down the fiber, and some of it escapes through the cladding. The light rays that escape represent a loss in light power. This is called Rayleigh scattering loss. Figure 20 graphically shows the relationship between wavelength and Rayleigh scat- tering loss. 10-3 Chromatic, or Wavelength, Dispersion Light-emitting diodes (LEDs) emit light containing many wavelengths. Each wavelength within the composite light signal travels at a different velocity when propagating through glass. Consequently, light rays that are simultaneously emitted from an LED and propagated down an optical fiber do not arrive at the far end of the fiber at the same time, resulting in an impairmentcalledchromaticdistortion(sometimescalledwavelengthdispersion).Chromatic distortion can be eliminated by using a monochromatic light source such as an injection laser diode (ILD). Chromatic distortion occurs only in fibers with a single mode of transmission. 10-4 Radiation Losses Radiation losses are caused mainly by small bends and kinks in the fiber. Essentially, there are two types of bends: microbends and constant-radius bends. Microbending occurs as a result of differences in the thermal contraction rates between the core and the cladding ma- terial.A microbend is a miniature bend or geometric imperfection along the axis of the fiber and represents a discontinuity in the fiber where Rayleigh scattering can occur. Mi- crobending losses generally contribute less than 20% of the total attenuation in a fiber. Optical Fiber Transmission Media 26
- 32. FIGURE 20 Rayleigh scattering loss as a function of wavelength Constant-radius bends are caused by excessive pressure and tension and generally occur when fibers are bent during handling or installation. 10-5 Modal Dispersion Modal dispersion (sometimes called pulse spreading) is caused by the difference in the propagation times of light rays that take different paths down a fiber. Obviously, modal dis- persion can occur only in multimode fibers. It can be reduced considerably by using graded- index fibers and almost entirely eliminated by using single-mode step-index fibers. Modal dispersion can cause a pulse of light energy to spread out in time as it propa- gates down a fiber. If the pulse spreading is sufficiently severe, one pulse may interfere with another. In multimode step-index fibers, a light ray propagating straight down the axis of the fiber takes the least amount of time to travel the length of the fiber.A light ray that strikes the core/cladding interface at the critical angle will undergo the largest number of internal reflections and, consequently, take the longest time to travel the length of the cable. For multimode propagation, dispersion is often expressed as a bandwidth length product (BLP) or bandwidth distance product (BDP). BLP indicates what signal frequen- cies can be propagated through a given distance of fiber cable and is expressed mathemat- ically as the product of distance and bandwidth (sometimes called linewidth). Bandwidth length products are often expressed in MHz km units. As the length of an optical cable increases, the bandwidth (and thus the bit rate) decreases in proportion. Example 4 For a 300-meter optical fiber cable with a BLP of 600 MHzkm, determine the bandwidth. Solution Figure 21 shows three light rays propagating down a multimode step-index optical fiber. The lowest-order mode (ray 1) travels in a path parallel to the axis of the fiber. The middle-order mode (ray 2) bounces several times at the interface before traveling the length B 2 GHz B 600 MHz km 0.3 km Optical Fiber Transmission Media 27
- 33. FIGURE 22 Light propagation down a single-mode step-index fiber FIGURE 23 Light propagation down a multimode graded-index fiber FIGURE 21 Light propagation down a multimode step-index fiber of the fiber. The highest-order mode (ray 3) makes many trips back and forth across the fiber as it propagates the entire length. It can be seen that ray 3 travels a considerably longer dis- tance than ray 1 over the length of the cable. Consequently, if the three rays of light were emitted into the fiber at the same time, each ray would reach the far end at a different time, resulting in a spreading out of the light energy with respect to time. This is called modal dispersion and results in a stretched pulse that is also reduced in amplitude at the output of the fiber. Figure 22 shows light rays propagating down a single-mode step-index cable. Be- cause the radial dimension of the fiber is sufficiently small, there is only a single transmis- sion path that all rays must follow as they propagate down the length of the fiber. Conse- quently, each ray of light travels the same distance in a given period of time, and modal dispersion is virtually eliminated. Figure 23 shows light propagating down a multimode graded-index fiber. Three rays are shown traveling in three different modes. Although the three rays travel differ- ent paths, they all take approximately the same amount of time to propagate the length of the fiber. This is because the refractive index decreases with distance from the center, and the velocity at which a ray travels is inversely proportional to the refractive index. Optical Fiber Transmission Media 28
- 34. FIGURE 24 Pulse-width dispersion in an optical fiber cable Consequently, the farther rays 2 and 3 travel from the center of the cable, the faster they propagate. Figure 24 shows the relative time/energy relationship of a pulse of light as it propa- gates down an optical fiber cable. From the figure, it can be seen that as the pulse propa- gates down the cable, the light rays that make up the pulse spread out in time, causing a cor- responding reduction in the pulse amplitude and stretching of the pulse width. This is called pulse spreading or pulse-width dispersion and causes errors in digital transmission. It can also be seen that as light energy from one pulse falls back in time, it will interfere with the next pulse, causing intersymbol interference. Figure 25a shows a unipolar return-to-zero (UPRZ) digital transmission. With UPRZ transmission (assuming a very narrow pulse), if light energy from pulse A were to fall back (spread) one bit time (tb), it would interfere with pulse B and change what was a logic 0 to a logic 1. Figure 25b shows a unipolar nonreturn-to-zero (UPNRZ) digital trans- mission where each pulse is equal to the bit time. With UPNRZ transmission, if energy from pulse A were to fall back one-half of a bit time, it would interfere with pulse B. Con- sequently, UPRZ transmissions can tolerate twice as much delay or spread as UPNRZ transmissions. The difference between the absolute delay times of the fastest and slowest rays of light propagating down a fiber of unit length is called the pulse-spreading constant ( t)and is gener- ally expressed in nanoseconds per kilometer (ns/km). The total pulse spread (T)is then equal to the pulse-spreading constant (t) times the total fiber length (L). Mathematically, T is T(ns) t(ns/km) L(km) (17) For UPRZ transmissions, the maximum data transmission rate in bits per second (bps) is expressed as (18) fb1bps2 1 ¢t L Optical Fiber Transmission Media 29
- 35. FIGURE 25 Pulse spreading of digital transmissions: (a) UPRZ; (b) UPNRZ and for UPNRZ transmissions, the maximum transmission rate is (19) Example 5 For an optical fiber 10 km long with a pulse-spreading constant of 5 ns/km, determine the maximum digital transmission rates for a. Return-to-zero. b. Nonreturn-to-zero transmissions. Solution a. Substituting into Equation 18 yields b. Substituting into Equation 19 yields The results indicate that the digital transmission rate possible for this optical fiber is twice as high (20 Mbps versus 10 Mbps) for UPRZ as for UPNRZ transmission. fb 1 12 5 nskm2 10 km 10 Mbps fb 1 5 nskm 10 km 20 Mbps fb1bps2 1 2¢t L Optical Fiber Transmission Media 30
- 36. FIGURE 26 Fiber alignment impairments: (a) lateral misalignment; (b) gap displacement; (c) angular misalign- ment; (d) surface finish 10-6 Coupling Losses Coupling losses are caused by imperfect physical connections. In fiber cables, coupling losses can occur at any of the following three types of optical junctions: light source-to-fiber connections, fiber-to-fiber connections, and fiber-to-photodetector connections. Junction losses are most often caused by one of the following alignment problems: lateral misalign- ment, gap misalignment, angular misalignment, and imperfect surface finishes. 10-6-1 Lateral displacement. Lateral displacement (misalignment) is shown in Figure 26a and is the lateral or axial displacement between two pieces of adjoining fiber ca- bles. The amount of loss can be from a couple tenths of a decibel to several decibels. This loss is generally negligible if the fiber axes are aligned to within 5% of the smaller fiber’s diameter. 10-6-2 Gap displacement (misalignment). Gap displacement (misalignment) is shown in Figure 26b and is sometimes called end separation. When splices are made in Optical Fiber Transmission Media 31
- 37. Optical Fiber Transmission Media 1 Yellow Green 2500°k 3400°k Tungsten lamp radiation spectrums for different temperatures 2000°k Red GaAs Eye response Ultraviolet wavelengths Infrared wavelengths Normalized human eye response 0.8 0.6 0.4 0.2 0 200 400 600 Wavelength (nanometers) 800 1000 1200 1400 Orange Blue FIGURE 27 Tungsten lamp radiation and human eye response optical fibers, the fibers should actually touch. The farther apart the fibers, the greater the loss of light. If two fibers are joined with a connector, the ends should not touch because the two ends rubbing against each other in the connector could cause damage to either or both fibers. 10-6-3 Angular displacement (misalignment). Angular displacement (misalign- ment) is shown in Figure 26c and is sometimes called angular displacement. If the angular displacement is less than 2°, the loss will typically be less than 0.5 dB. 10-6-4 Imperfect surface finish. Imperfect surface finish is shown in Figure 26d. The ends of the two adjoining fibers should be highly polished and fit together squarely. If the fiber ends are less than 3° off from perpendicular, the losses will typically be less than 0.5 dB. 11 LIGHT SOURCES The range of light frequencies detectable by the human eye occupies a very narrow segment of the total electromagnetic frequency spectrum. For example, blue light occupies the higher frequencies (shorter wavelengths) of visible light, and red hues occupy the lower fre- quencies (longer wavelengths). Figure 27 shows the light wavelength distribution produced from a tungsten lamp and the range of wavelengths perceivable by the human eye. As the figure shows, the human eye can detect only those lightwaves between approximately 380 nm and 780 nm. Furthermore, light consists of many shades of colors that are directly re- lated to the heat of the energy being radiated. Figure 27 also shows that more visible light is produced as the temperature of the lamp is increased. Light sources used for optical fiber systems must be at wavelengths efficiently propa- gated by the optical fiber. In addition, the range of wavelengths must be considered because the wider the range, the more likely the chance that chromatic dispersion will occur. Light 32
- 38. Optical Fiber Transmission Media Table 4 Semiconductor Material Wavelengths Material Wavelength (nm) AlGaInP 630–680 GaInP 670 GaAlAs 620–895 GaAs 904 InGaAs 980 InGaAsP 1100–1650 InGaAsSb 1700–4400 sources must also produce sufficient power to allow the light to propagate through the fiber without causing distortion in the cable itself or in the receiver. Lastly, light sources must be constructed so that their outputs can be efficiently coupled into and out of the optical cable. 12 OPTICAL SOURCES There are essentially only two types of practical light sources used to generate light for op- tical fiber communications systems: LEDs and ILDs. Both devices are constructed from semiconductor materials and have advantages and disadvantages. Standard LEDs have spectral widths of 30 nm to 50 nm, while injection lasers have spectral widths of only 1 nm to 3 nm (1 nm corresponds to a frequency of about 178 GHz). Therefore, a 1320-nm light source with a spectral linewidth of 0.0056 nm has a frequency bandwidth of approximately 1 GHz. Linewidth is the wavelength equivalent of bandwidth. Selection of one light-emitting device over the other is determined by system eco- nomic and performance requirements. The higher cost of laser diodes is offset by higher performance. LEDs typically have a lower cost and a corresponding lower performance. However, LEDs are typically more reliable. 12-1 LEDs AnLEDisap-njunctiondiode,usuallymadefromasemiconductormaterialsuchasaluminum- gallium-arsenide(AlGaAs)orgallium-arsenide-phosphide(GaAsP).LEDsemitlightbyspon- taneous emission—light is emitted as a result of the recombination of electrons and holes. When forward biased, minority carriers are injected across the p-n junction. Once across the junction, these minority carriers recombine with majority carriers and give up en- ergy in the form of light. This process is essentially the same as in a conventional semi- conductor diode except that in LEDs certain semiconductor materials and dopants are cho- sen such that the process is radiative; that is, a photon is produced. A photon is a quantum of electromagnetic wave energy. Photons are particles that travel at the speed of light but at rest have no mass. In conventional semiconductor diodes (germanium and silicon, for ex- ample), the process is primarily nonradiative, and no photons are generated. The energy gap of the material used to construct an LED determines the color of light it emits and whether the light emitted by it is visible to the human eye. To produce LEDs, semiconductors are formed from materials with atoms having ei- ther three or five valence electrons (known as Group III and Group IV atoms, respectively, because of their location in the periodic table of elements). To produce light wavelengths in the 800-nm range, LEDs are constructed from Group III atoms, such as gallium (Ga) and aluminum (Al), and a Group IV atom, such as arsenide (As). The junction formed is com- monly abbreviated GaAlAs for gallium-aluminum-arsenide. For longer wavelengths, gal- lium is combined with the Group III atom indium (In), and arsenide is combined with the GroupV atom phosphate (P), which forms a gallium-indium-arsenide-phosphate (GaInAsP) junction. Table 4 lists some of the common semiconductor materials used in LED con- struction and their respective output wavelengths. 33
- 39. Optical Fiber Transmission Media FIGURE 28 Homojunction LED structures: (a) silicon-doped gallium arsenide; (b) planar diffused 12-2 Homojunction LEDs A p-n junction made from two different mixtures of the same types of atoms is called a ho- mojunction structure. The simplest LED structures are homojunction and epitaxially grown, or they are single-diffused semiconductor devices, such as the two shown in Figure 28. Epit- axially grown LEDs are generally constructed of silicon-doped gallium-arsenide (Figure 28a). A typical wavelength of light emitted from this construction is 940 nm, and a typical output power is approximately 2 mW (3 dBm) at 100 mA of forward current. Light waves from homojunction sources do not produce a very useful light for an optical fiber. Light is emitted in all directions equally; therefore, only a small amount of the total light produced is coupled into the fiber. In addition, the ratio of electricity converted to light is very low. Homojunction devices are often called surface emitters. Planar diffused homojunction LEDs (Figure 28b) output approximately 500 μW at a wavelength of 900 nm. The primary disadvantage of homojunction LEDs is the nondirec- tionality of their light emission, which makes them a poor choice as a light source for op- tical fiber systems. 12-3 Heterojunction LEDs Heterojunction LEDs are made from a p-type semiconductor material of one set of atoms and an n-type semiconductor material from another set. Heterojunction devices are layered (usually two) such that the concentration effect is enhanced. This produces a device that confines the electron and hole carriers and the light to a much smaller area. The junction is generally manufactured on a substrate backing material and then sandwiched between metal contacts that are used to connect the device to a source of electricity. With heterojunction devices, light is emitted from the edge of the material and are therefore often called edge emitters. A planar heterojunction LED (Figure 29) is quite sim- ilar to the epitaxially grown LED except that the geometry is designed such that the forward current is concentrated to a very small area of the active layer. Heterojunction devices have the following advantages over homojunction devices: The increase in current density generates a more brilliant light spot. The smaller emitting area makes it easier to couple its emitted light into a fiber. The small effective area has a smaller capacitance, which allows the planar hetero- junction LED to be used at higher speeds. Figure 30 shows the typical electrical characteristics for a low-cost infrared light- emitting diode. Figure 30a shows the output power versus forward current. From the fig- ure, it can be seen that the output power varies linearly over a wide range of input current 34
- 40. Optical Fiber Transmission Media FIGURE 29 Planar heterojunction LED (0.5 mW [3 dBm] at 20 mA to 3.4 mW [5.3 dBm] at 140 mA). Figure 30b shows output power versus temperature. It can be seen that the output power varies inversely with tem- perature between a temperature range of 40°C to 80°C. Figure 30c shows relative output power in respect to output wavelength. For this particular example, the maximum output power is achieved at an output wavelength of 825 nm. 12-4 Burrus Etched-Well Surface-Emitting LED For the more practical applications, such as telecommunications, data rates in excess of 100 Mbps are required. For these applications, the etched-well LED was developed. Burrus and Dawson of Bell Laboratories developed the etched-well LED. It is a surface-emitting LED and is shown in Figure 31. The Burrus etched-well LED emits light in many directions. The etched well helps concentrate the emitted light to a very small area. Also, domed lenses can be placed over the emitting surface to direct the light into a smaller area. These devices are more efficient than the standard surface emitters, and they allow more power to be coupled into the optical fiber, but they are also more difficult and expensive to manufacture. 12-5 Edge-Emitting LED The edge-emitting LED, which was developed by RCA, is shown in Figure 32. These LEDs emit a more directional light pattern than do the surface-emitting LEDs. The construction is similar to the planar and Burrus diodes except that the emitting surface is a stripe rather than a confined circular area. The light is emitted from an active stripe and forms an ellip- tical beam. Surface-emitting LEDs are more commonly used than edge emitters because they emit more light. However, the coupling losses with surface emitters are greater, and they have narrower bandwidths. The radiant light power emitted from an LED is a linear function of the forward cur- rent passing through the device (Figure 33). It can also be seen that the optical output power of an LED is, in part, a function of the operating temperature. 12-6 ILD Lasers are constructed from many different materials, including gases, liquids, and solids, although the type of laser used most often for fiber-optic communications is the semicon- ductor laser. The ILD is similar to the LED. In fact, below a certain threshold current, an ILD acts similarly to an LED. Above the threshold current, an ILD oscillates; lasing occurs. As cur- rent passes through a forward-biased p-n junction diode, light is emitted by spontaneous emission at a frequency determined by the energy gap of the semiconductor material. When a particular current level is reached, the number of minority carriers and photons produced on either side of the p-n junction reaches a level where they begin to collide with already excited minority carriers. This causes an increase in the ionization energy level and makes the carriers unstable. When this happens, a typical carrier recombines with an opposite type 35
- 41. 3.5 3.0 2.5 1.5 0.5 0.0 0 20 40 60 80 Forward current (mA) Output power (mW) (a) 100 120 140 160 2.0 1.0 1.2 1.1 1.0 0.8 –60 –40 –20 0 20 Temperature, C Output power relative to 25°C (b) 40 60 80 100 0.9 1.0 0.8 0.6 0.2 0 700 750 800 Wavelength (nm) Relative power output (c) 850 900 0.4 FIGURE 30 Typical LED electrical characteristics: (a) output power-versus-forward current; (b) output power-versus-temperature; and (c) output power-versus-output wavelength 36
- 42. Optical Fiber Transmission Media FIGURE 31 Burrus etched-well surface-emitting LED FIGURE 32 Edge-emitting LED FIGURE 33 Output power versus forward current and operating temperature for an LED of carrier at an energy level that is above its normal before-collision value. In the process, two photons are created; one is stimulated by another. Essentially, a gain in the number of photons is realized. For this to happen, a large forward current that can provide many car- riers (holes and electrons) is required. The construction of an ILD is similar to that of an LED (Figure 34) except that the ends are highly polished. The mirrorlike ends trap the photons in the active region and, as they reflect back and forth, stimulate free electrons to recombine with holes at a higher- than-normal energy level. This process is called lasing. 37
- 43. Optical Fiber Transmission Media FIGURE 35 Output power versus forward current and temperature for an ILD FIGURE 34 Injection laser diode construction The radiant output light power of a typical ILD is shown in Figure 35. It can be seen that very little output power is realized until the threshold current is reached; then lasing occurs. After lasing begins, the optical output power increases dramatically, with small increases in drive current. It can also be seen that the magnitude of the op- tical output power of the ILD is more dependent on operating temperature than is the LED. Figure 36 shows the light radiation patterns typical of an LED and an ILD. Because light is radiated out the end of an ILD in a narrow concentrated beam, it has a more direct radiation pattern. ILDs have several advantages over LEDs and some disadvantages. Advantages in- clude the following: ILDs emit coherent (orderly) light, whereas LEDs emit incoherent (disorderly) light. Therefore, ILDs have a more direct radian pattern, making it easier to couple light emitted by the ILD into an optical fiber cable. This reduces the coupling losses and allows smaller fibers to be used. 38
- 44. Optical Fiber Transmission Media FIGURE 36 LED and ILD radiation patterns The radiant output power from an ILD is greater than that for an LED. A typical out- put power for an ILD is 5 mW (7 dBm) and only 0.5 mW (3 dBm) for LEDs. This allows ILDs to provide a higher drive power and to be used for systems that operate over longer distances. ILDs can be used at higher bit rates than LEDs. ILDs generate monochromatic light, which reduces chromatic or wavelength dispersion. Disadvantages include the following: ILDs are typically 10 times more expensive than LEDs. Because ILDs operate at higher powers, they typically have a much shorter lifetime than LEDs. ILDs are more temperature dependent than LEDs. 13 LIGHT DETECTORS There are two devices commonly used to detect light energy in fiber-optic communications receivers: PIN diodes and APDs. 13-1 PIN Diodes A PIN diode is a depletion-layer photodiode and is probably the most common device used as a light detector in fiber-optic communications systems. Figure 37 shows the basic con- struction of a PIN diode. A very lightly doped (almost pure or intrinsic) layer of n-type semi- conductor material is sandwiched between the junction of the two heavily doped n- and p- type contact areas. Light enters the device through a very small window and falls on the carrier-void intrinsic material. The intrinsic material is made thick enough so that most of the photons that enter the device are absorbed by this layer. Essentially, the PIN photodiode op- erates just the opposite of an LED. Most of the photons are absorbed by electrons in the va- lence band of the intrinsic material. When the photons are absorbed, they add sufficient en- ergy to generate carriers in the depletion region and allow current to flow through the device. 13-1-1 Photoelectric effect. Light entering through the window of a PIN diode is absorbed by the intrinsic material and adds enough energy to cause electronics to move from the valence band into the conduction band. The increase in the number of electrons that move into the conduction band is matched by an increase in the number of holes in the 39
- 45. Optical Fiber Transmission Media FIGURE 37 PIN photodiode construction valence band. To cause current to flow in a photodiode, light of sufficient energy must be absorbed to give valence electrons enough energy to jump the energy gap. The energy gap for silicon is 1.12 eV (electron volts). Mathematically, the operation is as follows: For silicon, the energy gap (Eg) equals 1.12 eV: 1 eV 1.6 1019 J Thus, the energy gap for silicon is and energy (E) hf (20) where h Planck’s constant 6.6256 1034 J/Hz f frequency (hertz) Rearranging and solving for f yields (21) For a silicon photodiode, Converting to wavelength yields 13-2 APDs Figure 38 shows the basic construction of an APD. An APD is a pipn structure. Light en- ters the diode and is absorbed by the thin, heavily doped n-layer. A high electric field in- tensity developed across the i-p-n junction by reverse bias causes impact ionization to oc- cur. During impact ionization, a carrier can gain sufficient energy to ionize other bound electrons. These ionized carriers, in turn, cause more ionizations to occur. The process con- tinues as in an avalanche and is, effectively, equivalent to an internal gain or carrier multi- plication. Consequently, APDs are more sensitive than PIN diodes and require less addi- tional amplification. The disadvantages of APDs are relatively long transit times and additional internally generated noise due to the avalanche multiplication factor. λ c f 3 108 ms 2.705 1014 Hz 1109 nmcycle f 1.792 1019 J 6.6256 1034 JHz 2.705 1014 Hz f E h Eg 11.12 eV2¢1.6 1019 J eV ≤ 1.792 1019 J 40
- 46. Optical Fiber Transmission Media FIGURE 39 Spectral response curve FIGURE 38 Avalanche photo-diode construction 13-3 Characteristics of Light Detectors The most important characteristics of light detectors are the following: 1. Responsivity. A measure of the conversion efficiency of a photodetector. It is the ratio of the output current of a photodiode to the input optical power and has the unit of amperes per watt. Responsivity is generally given for a particular wave- length or frequency. 2. Dark current. The leakage current that flows through a photodiode with no light input. Thermally generated carriers in the diode cause dark current. 3. Transit time. The time it takes a light-induced carrier to travel across the depletion region of a semiconductor. This parameter determines the maximum bit rate pos- sible with a particular photodiode. 4. Spectral response. The range of wavelength values that a given photodiode will respond. Generally, relative spectral response is graphed as a function of wave- length or frequency, as shown in Figure 39. 5. Light sensitivity. The minimum optical power a light detector can receive and still produce a usable electrical output signal. Light sensitivity is generally given for a particular wavelength in either dBm or dBμ. 14 LASERS Laser is an acronym for light amplification stimulated by the emission of radiation. Laser technology deals with the concentration of light into a very small, powerful beam. The acronym was chosen when technology shifted from microwaves to light waves. Basically, there are four types of lasers: gas, liquid, solid, and semiconductor. 41
- 47. Optical Fiber Transmission Media The first laser was developed by Theodore H. Maiman, a scientist who worked for Hughes Aircraft Company in California. Maiman directed a beam of light into ruby crys- tals with a xenon flashlamp and measured emitted radiation from the ruby. He discovered that when the emitted radiation increased beyond threshold, it caused emitted radiation to become extremely intense and highly directional. Uranium lasers were developed in 1960 along with other rare-earth materials.Also in 1960,A. Javin of Bell Laboratories developed the helium laser. Semiconductor lasers (injection laser diodes) were manufactured in 1962 by General Electric, IBM, and Lincoln Laboratories. 14-1 Laser Types Basically, there are four types of lasers: gas, liquid, solid, and semiconductor. 1. Gas lasers. Gas lasers use a mixture of helium and neon enclosed in a glass tube. A flow of coherent (one frequency) light waves is emitted through the output cou- pler when an electric current is discharged into the gas. The continuous light-wave output is monochromatic (one color). 2. Liquid lasers. Liquid lasers use organic dyes enclosed in a glass tube for an active medium. Dye is circulated into the tube with a pump.A powerful pulse of light ex- cites the organic dye. 3. Solid lasers. Solid lasers use a solid, cylindrical crystal, such as ruby, for the active medium. Each end of the ruby is polished and parallel. The ruby is excited by a tung- sten lamp tied to an ac power supply. The output from the laser is a continuous wave. 4. Semiconductor lasers. Semiconductor lasers are made from semiconductor p-n junctions and are commonly called ILDs. The excitation mechanism is a dc power supply that controls the amount of current to the active medium. The output light from an ILD is easily modulated, making it very useful in many electronic com- munications applications. 14-2 Laser Characteristics All types of lasers have several common characteristics. They all use (1) an active material to convert energy into laser light, (2) a pumping source to provide power or energy, (3) op- tics to direct the beam through the active material to be amplified, (4) optics to direct the beam into a narrow powerful cone of divergence, (5) a feedback mechanism to provide con- tinuous operation, and (6) an output coupler to transmit power out of the laser. The radiation of a laser is extremely intense and directional. When focused into a fine hairlike beam, it can concentrate all its power into the narrow beam. If the beam of light were allowed to diverge, it would lose most of its power. 14-3 Laser Construction Figure 40 shows the construction of a basic laser. A power source is connected to a flash- tube that is coiled around a glass tube that holds the active medium. One end of the glass tube is a polished mirror face for 100% internal reflection. The flashtube is energized by a trigger pulse and produces a high-level burst of light (similar to a flashbulb). The flash causes the chromium atoms within the active crystalline structure to become excited. The process of pumping raises the level of the chromium atoms from ground state to an excited energy state. The ions then decay, falling to an intermediate energy level. When the pop- ulation of ions in the intermediate level is greater than the ground state, a population in- version occurs. The population inversion causes laser action (lasing) to occur. After a pe- riod of time, the excited chromium atoms will fall to the ground energy level. At this time, photons are emitted. A photon is a packet of radiant energy. The emitted photons strike atoms and two other photons are emitted (hence the term “stimulated emission”). The fre- quency of the energy determines the strength of the photons; higher frequencies cause greater-strength photons. 42
- 48. Optical Fiber Transmission Media FIGURE 40 Laser construction 15 OPTICAL FIBER SYSTEM LINK BUDGET As with any communications system, optical fiber systems consist of a source and a desti- nation that are separated by numerous components and devices that introduce various amounts of loss or gain to the signal as it propagates through the system. Figure 41 shows two typical optical fiber communications system configurations. Figure 41a shows a re- peaterless system where the source and destination are interconnected through one or more sections of optical cable. With a repeaterless system, there are no amplifiers or regenerators between the source and destination. Figure 41b shows an optical fiber system that includes a repeater that either amplifies or regenerates the signal. Repeatered systems are obviously used when the source and des- tination are separated by great distances. Link budgets are generally calculated between a light source and a light detector; therefore, for our example, we look at a link budget for a repeaterless system. A repeater- less system consists of a light source, such as an LED or ILD, and a light detector, such as an APD connected by optical fiber and connectors. Therefore, the link budget consists of a light power source, a light detector, and various cable and connector losses. Losses typical to optical fiber links include the following: 1. Cable losses. Cable losses depend on cable length, material, and material purity. They are generally given in dB/km and can vary between a few tenths of a dB to several dB per kilometer. 2. Connector losses. Mechanical connectors are sometimes used to connect two sec- tions of cable. If the mechanical connection is not perfect, light energy can escape, resulting in a reduction in optical power. Connector losses typically vary between a few tenths of a dB to as much as 2 dB for each connector. 3. Source-to-cable interface loss. The mechanical interface used to house the light source and attach it to the cable is seldom perfect. Therefore, a small percentage of optical power is not coupled into the cable, representing a power loss to the sys- tem of several tenths of a dB. 4. Cable-to-light detector interface loss. The mechanical interface used to house the light detector and attach it to the cable is also not perfect and, therefore, prevents a small percentage of the power leaving the cable from entering the light detector. This, of course, represents a loss to the system usually of a few tenths of a dB. 5. Splicing loss. If more than one continuous section of cable is required, cable sec- tions can be fused together (spliced). Because the splices are not perfect, losses ranging from a couple tenths of a dB to several dB can be introduced to the signal. 43
- 49. Signal source Optical transmitter (LED or ILD) Optical receiver (APD) Signal destination Fiber cable Fiber cable (a) Signal source Optical transmitter (LED or ILD) Optical receiver (APD) Repeater (Amplifier or regenerator) Signal destination Fiber cable (b) FIGURE 41 Optical fiber communications systems: (a) without repeaters; (b) with repeaters 6. Cable bends. When an optical cable is bent at too large an angle, the internal char- acteristics of the cable can change dramatically. If the changes are severe, total re- flections for some of the light rays may no longer be achieved, resulting in refrac- tion. Light refracted at the core/cladding interface enters the cladding, resulting in a net loss to the signal of a few tenths of a dB to several dB. As with any link or system budget, the useful power available in the receiver depends on transmit power and link losses. Mathematically, receive power is represented as Pr Pt losses (22) where Pr power received (dBm) Pt power transmitted (dBm) losses sum of all losses (dB) Example 6 Determine the optical power received in dBm and watts for a 20-km optical fiber link with the fol- lowing parameters: LED output power of 30 mW Four 5-km sections of optical cable each with a loss of 0.5 dB/km Three cable-to-cable connectors with a loss of 2 dB each No cable splices Light source-to-fiber interface loss of 1.9 dB Fiber-to-light detector loss of 2.1 dB No losses due to cable bends Solution The LED output power is converted to dBm using Equation 6: 14.8 dBm Pout 10 log 30 mW 1 mW Optical Fiber Transmission Media 44
- 50. The cable loss is simply the product of the total cable length in km and the loss in dB/km. Four 5-km sections of cable is a total cable length of 20 km; therefore, total cable loss 20 km 0.5 dB/km 10 dB Cable connector loss is simply the product of the loss in dB per connector and the number of con- nectors. The maximum number of connectors is always one less than the number of sections of cable. Four sections of cable would then require three connectors; therefore, total connector loss 3 connectors 2 dB/connector 6 dB The light source-to-cable and cable-to-light detector losses were given as 1.9 dB and 2.1 dB, respec- tively. Therefore, total loss cable loss connector loss light source-to-cable loss cable-to-light detector loss 10 dB 6 dB 1.9 dB 2.1 dB 20 dB The receive power is determined by substituting into Equation 22: Pr 14.8 dBm 20 dB 5.2 dBm 0.302 mW QUESTIONS 1. Define a fiber-optic system. 2. What is the relationship between information capacity and bandwidth? 3. What development in 1951 was a substantial breakthrough in the field of fiber optics? In 1960? In 1970? 4. Contrast the advantages and disadvantages of fiber-optic cables and metallic cables. 5. Outline the primary building blocks of a fiber-optic system. 6. Contrast glass and plastic fiber cables. 7. Briefly describe the construction of a fiber-optic cable. 8. Define the following terms: velocity of propagation, refraction, and refractive index. 9. State Snell’s law for refraction and outline its significance in fiber-optic cables. 10. Define critical angle. 11. Describe what is meant by mode of operation; by index profile. 12. Describe a step-index fiber cable; a graded-index cable. 13. Contrast the advantages and disadvantages of step-index, graded-index, single-mode, and multi- mode propagation. 14. Why is single-mode propagation impossible with graded-index fibers? 15. Describe the source-to-fiber aperture. 16. What are the acceptance angle and the acceptance cone for a fiber cable? 17. Define numerical aperture. 18. List and briefly describe the losses associated with fiber cables. 19. What is pulse spreading? 20. Define pulse spreading constant. 21. List and briefly describe the various coupling losses. 22. Briefly describe the operation of a light-emitting diode. 23. What are the two primary types of LEDs? 24. Briefly describe the operation of an injection laser diode. 25. What is lasing? 26. Contrast the advantages and disadvantages of ILDs and LEDs. 27. Briefly describe the function of a photodiode. 28. Describe the photoelectric effect. Optical Fiber Transmission Media 45
- 51. 29. Explain the difference between a PIN diode and an APD. 30. List and describe the primary characteristics of light detectors. PROBLEMS 1. Determine the wavelengths in nanometers and angstroms for the following light frequencies: a. 3.45 1014 Hz b. 3.62 1014 Hz c. 3.21 1014 Hz 2. Determine the light frequency for the following wavelengths: a. 670 nm b. 7800 Å c. 710 nm 3. For a glass (n 1.5)/quartz (n 1.38) interface and an angle of incidence of 35°, determine the angle of refraction. 4. Determine the critical angle for the fiber described in problem 3. 5. Determine the acceptance angle for the cable described in problem 3. 6. Determine the numerical aperture for the cable described in problem 3. 7. Determine the maximum bit rate for RZ and NRZ encoding for the following pulse-spreading con- stants and cable lengths: a. t 10 ns/m, L 100 m b. t 20 ns/m, L 1000 m c. t 2000 ns/km, L 2 km 8. Determine the lowest light frequency that can be detected by a photodiode with an energy gap 1.2 eV. 9. Determine the wavelengths in nanometers and angstroms for the following light frequencies: a. 3.8 1014 Hz b. 3.2 1014 Hz c. 3.5 1014 Hz 10. Determine the light frequencies for the following wavelengths: a. 650 nm b. 7200 Å c. 690 nm 11. For a glass (n 1.5)/quartz (n 1.41) interface and an angle of incidence of 38°, determine the angle of refraction. 12. Determine the critical angle for the fiber described in problem 11. 13. Determine the acceptance angle for the cable described in problem 11. 14. Determine the numerical aperture for the cable described in problem 11. 15. Determine the maximum bit rate for RZ and NRZ encoding for the following pulse-spreading con- stants and cable lengths: a. t 14 ns/m, L 200 m b. t 10 ns/m, L 50 m c. t 20 ns/m, L 200 m 16. Determine the lowest light frequency that can be detected by a photodiode with an energy gap 1.25 eV. 17. Determine the optical power received in dBm and watts for a 24-km optical fiber link with the following parameters: LED output power of 20 mW Six 4-km sections of optical cable each with a loss of 0.6 dB/km Three cable-to-cable connectors with a loss of 2.1 dB each No cable splices Light source-to-fiber interface loss of 2.2 dB Fiber-to-light detector loss of 1.8 dB No losses due to cable bends Optical Fiber Transmission Media 46
- 52. ANSWERS TO SELECTED PROBLEMS 1. a. 869 nm, 8690 A° b. 828 nm, 8280 A° c. 935 nm, 9350 A° 3. 38.57° 5. 56° 7. a. RZ 1 Mbps, NRZ 500 kbps b. RZ 50 kbps, NRZ 25 kbps c. RZ 250 kbps, NRZ 125 kbps 9. a. 789 nm, 7890 A° b. 937 nm, 9370 A° c. 857 nm, 8570 A° 11. 42° 13. 36° 15. a. RZ 357 kbps, NRZ 179 kbps b. RZ 2 Mbps, NRZ 1 Mbps c. RZ 250 kbps, NRZ 125 kbps Optical Fiber Transmission Media 47
- 53. 48
- 54. CHAPTER OUTLINE 1 Introduction 2 Information Capacity, Bits, Bit Rate, Baud, and M-ary Encoding 3 Amplitude-Shift Keying 4 Frequency-Shift Keying 5 Phase-Shift Keying 6 Quadrature-Amplitude Modulation 7 Bandwidth Efficiency 8 Carrier Recovery 9 Clock Recovery 10 Differential Phase-Shift Keying 11 Trellis Code Modulation 12 Probability of Error and Bit Error Rate 13 Error Performance OBJECTIVES ■ Define electronic communications ■ Define digital modulation and digital radio ■ Define digital communications ■ Define information capacity ■ Define bit, bit rate, baud, and minimum bandwidth ■ Explain Shannon’s limit for information capacity ■ Explain M-ary encoding ■ Define and describe digital amplitude modulation ■ Define and describe frequency-shift keying ■ Describe continuous-phase frequency-shift keying ■ Define phase-shift keying ■ Explain binary phase-shift keying ■ Explain quaternary phase-shift keying Digital Modulation ■ Describe 8- and 16-PSK ■ Describe quadrature-amplitude modulation ■ Explain 8-QAM ■ Explain 16-QAM ■ Define bandwidth efficiency ■ Explain carrier recovery ■ Explain clock recovery ■ Define and describe differential phase-shift keying ■ Define and explain trellis-code modulation ■ Define probability of error and bit error rate ■ Develop error performance equations for FSK, PSK, and QAM From Chapter 2 of Advanced Electronic Communications Systems, Sixth Edition. Wayne Tomasi. Copyright © 2004 by Pearson Education, Inc. Published by Pearson Prentice Hall. All rights reserved. 49
- 55. 1 INTRODUCTION In essence, electronic communications is the transmission, reception, and processing of in- formation with the use of electronic circuits. Information is defined as knowledge or intel- ligence that is communicated (i.e., transmitted or received) between two or more points. Digital modulation is the transmittal of digitally modulated analog signals (carriers) be- tween two or more points in a communications system. Digital modulation is sometimes called digital radio because digitally modulated signals can be propagated through Earth’s atmosphere and used in wireless communications systems. Traditional electronic commu- nications systems that use conventional analog modulation, such as amplitude modulation (AM), frequency modulation (FM), and phase modulation (PM), are rapidly being replaced with more modern digital moduluation systems that offer several outstanding advantages over traditional analog systems, such as ease of processing, ease of multiplexing, and noise immunity. Digital communications is a rather ambiguous term that could have entirely different meanings to different people. In the context of this text, digital communications include systems where relatively high-frequency analog carriers are modulated by relatively low- frequency digital information signals (digital radio) and systems involving the transmission of digital pulses (digital transmission). Digital transmission systems transport information in digital form and, therefore, require a physical facility between the transmitter and re- ceiver, such as a metallic wire pair, a coaxial cable, or an optical fiber cable. In digital ra- dio systems, the carrier facility could be a physical cable, or it could be free space. The property that distinguishes digital radio systems from conventional analog- modulation communications systems is the nature of the modulating signal. Both analog and digital modulation systems use analog carriers to transport the information through the system. However, with analog modulation systems, the information signal is also analog, whereas with digital modulation, the information signal is digital, which could be computer- generated data or digitally encoded analog signals. Referring to Equation 1, if the information signal is digital and the amplitude (V) of the carrier is varied proportional to the information signal, a digitally modulated signal called amplitude shift keying (ASK) is produced. If the frequency (f) is varied proportional to the information signal, frequency shift keying (FSK) is produced, and if the phase of the carrier (θ) is varied proportional to the information signal, phase shift keying (PSK) is pro- duced. If both the amplitude and the phase are varied proportional to the information sig- nal, quadrature amplitude modulation (QAM) results. ASK, FSK, PSK, and QAM are all forms of digital modulation: (1) Digital modulation is ideally suited to a multitude of communications applications, including both cable and wireless systems. Applications include the following: (1) rela- tively low-speed voice-band data communications modems, such as those found in most personal computers; (2) high-speed data transmission systems, such as broadband digital subscriber lines (DSL); (3) digital microwave and satellite communications systems; and (4) cellular telephone Personal Communications Systems (PCS). Figure 1 shows a simplified block diagram for a digital modulation system. In the transmitter, the precoder performs level conversion and then encodes the incoming data into groups of bits that modulate an analog carrier. The modulated carrier is shaped (fil- v(t) = V sin (2p . ƒt + U) ASK FSK QAM PSK Digital Modulation 50
- 56. FIGURE 1 Simplified block diagram of a digital radio system tered), amplified, and then transmitted through the transmission medium to the receiver. The transmission medium can be a metallic cable, optical fiber cable, Earth’s atmosphere, or a combination of two or more types of transmission systems. In the receiver, the in- coming signals are filtered, amplified, and then applied to the demodulator and decoder circuits, which extracts the original source information from the modulated carrier. The clock and carrier recovery circuits recover the analog carrier and digital timing (clock) signals from the incoming modulated wave since they are necessary to perform the de- modulation process. 2 INFORMATION CAPACITY, BITS, BIT RATE, BAUD, AND M-ARY ENCODING 2-1 Information Capacity, Bits, and Bit Rate Information theory is a highly theoretical study of the efficient use of bandwidth to propa- gate information through electronic communications systems. Information theory can be used to determine the information capacity of a data communications system. Information capacity is a measure of how much information can be propagated through a communica- tions system and is a function of bandwidth and transmission time. Information capacity represents the number of independent symbols that can be car- ried through a system in a given unit of time. The most basic digital symbol used to repre- sent information is the binary digit, or bit. Therefore, it is often convenient to express the information capacity of a system as a bit rate. Bit rate is simply the number of bits trans- mitted during one second and is expressed in bits per second (bps). In 1928, R. Hartley of Bell Telephone Laboratories developed a useful relationship among bandwidth, transmission time, and information capacity. Simply stated, Hartley’s law is I ∝ B t (2) where I information capacity (bits per second) B bandwidth (hertz) t transmission time (seconds) Digital Modulation 51
- 57. From Equation 2, it can be seen that information capacity is a linear function of band- width and transmission time and is directly proportional to both. If either the bandwidth or the transmission time changes, a directly proportional change occurs in the information ca- pacity. In 1948, mathematician Claude E. Shannon (also of Bell Telephone Laboratories) published a paper in the Bell System Technical Journal relating the information capacity of a communications channel to bandwidth and signal-to-noise ratio. The higher the signal- to-noise ratio, the better the performance and the higher the information capacity. Mathe- matically stated, the Shannon limit for information capacity is (3) or (4) where I information capacity (bps) B bandwidth (hertz) signal-to-noise power ratio (unitless) For a standard telephone circuit with a signal-to-noise power ratio of 1000 (30 dB) and a bandwidth of 2.7 kHz, the Shannon limit for information capacity is I (3.32)(2700) log10 (1 1000) 26.9 kbps Shannon’s formula is often misunderstood. The results of the preceding example indi- cate that 26.9 kbps can be propagated through a 2.7-kHz communications channel. This may be true, but it cannot be done with a binary system. To achieve an information transmission rate of 26.9 kbps through a 2.7-kHz channel, each symbol transmitted must contain more than one bit. 2-2 M-ary Encoding M-ary is a term derived from the word binary. M simply represents a digit that corresponds to the number of conditions, levels, or combinations possible for a given number of binary variables. It is often advantageous to encode at a level higher than binary (sometimes re- ferred to as beyond binary or higher-than-binary encoding) where there are more than two conditions possible. For example, a digital signal with four possible conditions (voltage lev- els, frequencies, phases, and so on) is an M-ary system where M 4. If there are eight pos- sible conditions, M 8 and so forth. The number of bits necessary to produce a given num- ber of conditions is expressed mathematically as N log2 M (5) where N number of bits necessary M number of conditions, levels, or combinations possible with N bits Equation 5 can be simplified and rearranged to express the number of conditions possible with N bits as 2N M (6) For example, with one bit, only 21 2 conditions are possible. With two bits, 22 4 con- ditions are possible, with three bits, 23 8 conditions are possible, and so on. S N I 3.32B log10 ¢1 S N ≤ I B log2¢1 S N ≤ Digital Modulation 52
- 58. 2-3 Baud and Minimum Bandwidth Baud is a term that is often misunderstood and commonly confused with bit rate (bps). Bit rate refers to the rate of change of a digital information signal, which is usually binary. Baud, like bit rate, is also a rate of change; however, baud refers to the rate of change of a signal on the transmission medium after encoding and modulation have occurred. Hence, baud is a unit of transmission rate, modulation rate, or symbol rate and, therefore, the terms symbols per second and baud are often used interchangeably. Mathematically, baud is the reciprocal of the time of one output signaling element, and a signaling element may repre- sent several information bits. Baud is expressed as (7) where baud symbol rate (baud per second) ts time of one signaling element (seconds) A signaling element is sometimes called a symbol and could be encoded as a change in the amplitude, frequency, or phase. For example, binary signals are generally encoded and transmitted one bit at a time in the form of discrete voltage levels representing logic 1s (highs) and logic 0s (lows). A baud is also transmitted one at a time; however, a baud may represent more than one information bit. Thus, the baud of a data communications system may be considerably less than the bit rate. In binary systems (such as binary FSK and bi- nary PSK), baud and bits per second are equal. However, in higher-level systems (such as QPSK and 8-PSK), bps is always greater than baud. According to H. Nyquist, binary digital signals can be propagated through an ideal noiseless transmission medium at a rate equal to two times the bandwidth of the medium. The minimum theoretical bandwidth necessary to propagate a signal is called the minimum Nyquist bandwidth or sometimes the minimum Nyquist frequency. Thus, fb 2B, where fb is the bit rate in bps and B is the ideal Nyquist bandwidth. The actual bandwidth necessary to propagate a given bit rate depends on several factors, including the type of encoding and modulation used, the types of filters used, system noise, and desired error performance. The ideal bandwidth is generally used for comparison purposes only. Therelationshipbetweenbandwidthandbitratealsoappliestotheoppositesituation.For a given bandwidth (B), the highest theoretical bit rate is 2B. For example, a standard telephone circuit has a bandwidth of approximately 2700 Hz, which has the capacity to propagate 5400 bps through it. However, if more than two levels are used for signaling (higher-than-binary encoding), more than one bit may be transmitted at a time, and it is possible to propagate a bit rate that exceeds 2B. Using multilevel signaling, the Nyquist formulation for channel capacity is fb 2B log2 M (8) where fb channel capacity (bps) B minimum Nyquist bandwidth (hertz) M number of discrete signal or voltage levels Equation 8 can be rearranged to solve for the minimum bandwidth necessary to pass M-ary digitally modulated carriers (9) If N is substituted for log2 M, Equation 9 reduces to (10) where N is the number of bits encoded into each signaling element. B fb N B ¢ fb log2 M ≤ baud 1 ts Digital Modulation 53
- 59. If information bits are encoded (grouped) and then converted to signals with more than two levels, transmission rates in excess of 2B are possible, as will be seen in subse- quent sections of this chapter. In addition, since baud is the encoded rate of change, it also equals the bit rate divided by the number of bits encoded into one signaling element. Thus, (11) By comparing Equation 10 with Equation 11, it can be seen that with digital modu- lation, the baud and the ideal minimum Nyquist bandwidth have the same value and are equal to the bit rate divided by the number of bits encoded. This statement holds true for all forms of digital modulation except frequency-shift keying. 3 AMPLITUDE-SHIFT KEYING The simplest digital modulation technique is amplitude-shift keying (ASK), where a binary information signal directly modulates the amplitude of an analog carrier. ASK is similar to standard amplitude modulation except there are only two output amplitudes possible. Amplitude-shift keying is sometimes called digital amplitude modulation (DAM). Mathe- matically, amplitude-shift keying is (12) where vask(t) amplitude-shift keying wave vm(t) digital information (modulating) signal (volts) A/2 unmodulated carrier amplitude (volts) ωc analog carrier radian frequency (radians per second, 2π fct) In Equation 12, the modulating signal (vm[t]) is a normalized binary waveform, where 1 V logic 1 and 1 V logic 0. Therefore, for a logic 1 input, vm(t) 1 V, Equation 12 re- duces to and for a logic 0 input, vm(t) 1 V, Equation 12 reduces to 0 Thus, the modulated wave vask(t), is either A cos(ωct) or 0. Hence, the carrier is either “on” or “off,” which is why amplitude-shift keying is sometimes referred to as on-off keying (OOK). Figure 2 shows the input and output waveforms from an ASK modulator. From the figure, it can be seen that for every change in the input binary data stream, there is one change in the ASK waveform, and the time of one bit (tb) equals the time of one analog sig- naling element (ts). It is also important to note that for the entire time the binary input is high, the output is a constant-amplitude, constant-frequency signal, and for the entire time the bi- nary input is low, the carrier is off. The bit time is the reciprocal of the bit rate and the time of one signaling element is the reciprocal of the baud. Therefore, the rate of change of the v1ask21t2 31 14B A 2 cos1ct2R A cos1ct2 v1ask21t2 31 14B A 2 cos1ct2R v1ask21t2 31 vm1t2 4B A 2 cos1ct2R baud fb N Digital Modulation 54
- 60. (a) Binary input DAM output (b) FIGURE 2 Digital amplitude modula- tion: (a) input binary; (b) output DAM waveform ASK waveform (baud) is the same as the rate of change of the binary input (bps); thus, the bit rate equals the baud. With ASK, the bit rate is also equal to the minimum Nyquist band- width. This can be verified by substituting into Equations 10 and 11 and setting N to 1: Example 1 Determine the baud and minimum bandwidth necessary to pass a 10 kbps binary signal using ampli- tude shift keying. Solution For ASK, N 1, and the baud and minimum bandwidth are determined from Equations 11 and 10, respectively: The use of amplitude-modulated analog carriers to transport digital information is a relatively low-quality, low-cost type of digital modulation and, therefore, is seldom used except for very low- speed telemetry circuits. 4 FREQUENCY-SHIFT KEYING Frequency-shift keying (FSK) is another relatively simple, low-performance type of digital modulation. FSK is a form of constant-amplitude angle modulation similar to standard fre- quency modulation (FM) except the modulating signal is a binary signal that varies between two discrete voltage levels rather than a continuously changing analog waveform. Conse- quently, FSK is sometimes called binary FSK (BFSK). The general expression for FSK is vfsk(t) Vc cos{2π[fc vm(t) f]t} (13) where vfsk(t) binary FSK waveform Vc peak analog carrier amplitude (volts) fc analog carrier center frequency (hertz) f peak change (shift) in the analog carrier frequency (hertz) vm(t) binary input (modulating) signal (volts) From Equation 13, it can be seen that the peak shift in the carrier frequency (f) is proportional to the amplitude of the binary input signal (vm[t]), and the direction of the shift baud 10,000 1 10,000 B 10,000 1 10,000 B fb 1 fb baud fb 1 fb Digital Modulation 55
- 61. –Δƒ +Δƒ ƒc ƒs ƒm Binary input signal Logic 1 Logic 0 FIGURE 3 FSK in the frequency domain is determined by the polarity. The modulating signal is a normalized binary waveform where a logic 1 1 V and a logic 0 1 V. Thus, for a logic 1 input, vm(t) 1, Equa- tion 13 can be rewritten as vfsk(t) Vc cos[2π(fc f)t] For a logic 0 input, vm(t) 1, Equation 13 becomes vfsk(t) Vc cos[2π(fc f)t] With binary FSK, the carrier center frequency (fc) is shifted (deviated) up and down in the frequency domain by the binary input signal as shown in Figure 3. As the binary in- put signal changes from a logic 0 to a logic 1 and vice versa, the output frequency shifts be- tween two frequencies: a mark, or logic 1 frequency (fm), and a space, or logic 0 frequency (fs). The mark and space frequencies are separated from the carrier frequency by the peak frequency deviation (f) and from each other by 2 f. With FSK, frequency deviation is defined as the difference between either the mark or space frequency and the center frequency, or half the difference between the mark and space frequencies. Frequency deviation is illustrated in Figure 3 and expressed mathemat- ically as (14) where f frequency deviation (hertz) |fm fs| absolute difference between the mark and space frequencies (hertz) Figure 4a shows in the time domain the binary input to an FSK modulator and the cor- responding FSK output.As the figure shows, when the binary input (fb) changes from a logic 1 to a logic 0 and vice versa, the FSK output frequency shifts from a mark (fm) to a space (fs) frequency and vice versa. In Figure 4a, the mark frequency is the higher frequency (fc f), and the space frequency is the lower frequency (fc f), although this relationship could be just the opposite. Figure 4b shows the truth table for a binary FSK modulator. The truth table shows the input and output possibilities for a given digital modulation scheme. 4-1 FSK Bit Rate, Baud, and Bandwidth In Figure 4a, it can be seen that the time of one bit (tb) is the same as the time the FSK out- put is a mark of space frequency (ts). Thus, the bit time equals the time of an FSK signal- ing element, and the bit rate equals the baud. ¢f 冟 fm fs冟 2 Digital Modulation 56
- 62. ts tb 0 Binary input (a) Analog output ƒm, mark frequency; ƒs space frequency 1 0 1 0 1 0 1 0 1 0 ƒs ƒm ƒs ƒm ƒs ƒm ƒs ƒm ƒs ƒm ƒs FIGURE 4 FSK in the time domain: (a) waveform; (b) truth table The baud for binary FSK can also be determined by substituting N 1 in Equa- tion 11: FSK is the exception to the rule for digital modulation, as the minimum bandwidth is not determined from Equation 10. The minimum bandwidth for FSK is given as B |(fs fb) (fm fb)| |fs fm| 2 fb and since |fs fm| equals 2f, the minimum bandwidth can be approximated as B 2(f fb) (15) where B minimum Nyquist bandwidth (hertz) f frequency deviation (|fm fs|) (hertz) fb input bit rate (bps) Note how closely Equation 15 resembles Carson’s rule for determining the approxi- mate bandwidth for an FM wave. The only difference in the two equations is that, for FSK, the bit rate (fb) is substituted for the modulating-signal frequency (fm). Example 2 Determine (a) the peak frequency deviation, (b) minimum bandwidth, and (c) baud for a binary FSK signal with a mark frequency of 49 kHz, a space frequency of 51 kHz, and an input bit rate of 2 kbps. Solution a. The peak frequency deviation is determined from Equation 14: 1 kHz b. The minimum bandwidth is determined from Equation 15: B 2(1000 2000) 6 kHz c. For FSK, N 1, and the baud is determined from Equation 11 as baud 2000 1 2000 ¢f 冟49kHz 51kHz冟 2 baud fb 1 fb Digital Modulation binary frequency input output 0 space (fs) 1 mark (fm) (b) 57
- 63. FIGURE 5 FSK modulator, tb, time of one bit = 1/fb; fm, mark frequency; fs, space frequency; T1, period of shortest cycle; 1/T1, fundamental frequency of binary square wave; fb, input bit rate (bps) Bessel functions can also be used to determine the approximate bandwidth for an FSK wave.As shown in Figure 5, the fastest rate of change (highest fundamental frequency) in a nonreturn-to-zero (NRZ) binary signal occurs when alternating 1s and 0s are occurring (i.e., a square wave). Since it takes a high and a low to produce a cycle, the highest funda- mental frequency present in a square wave equals the repetition rate of the square wave, which with a binary signal is equal to half the bit rate. Therefore, (16) where fa highest fundamental frequency of the binary input signal (hertz) fb input bit rate (bps) The formula used for modulation index in FM is also valid for FSK; thus, (17) where h FM modulation index called the h-factor in FSK fa fundamental frequency of the binary modulating signal (hertz) f peak frequency deviation (hertz) The worst-case modulation index (deviation ratio) is that which yields the widest band- width. The worst-case or widest bandwidth occurs when both the frequency deviation and the modulating-signal frequency are at their maximum values. As described earlier, the peak frequency deviation in FSK is constant and always at its maximum value, and the highest fundamental frequency is equal to half the incoming bit rate. Thus, or (18) h 冟fm fs冟 fb h 冟fm fs冟 2 fb 2 1unitless2 h ¢f fa 1unitless2 fa fb 2 Digital Modulation 58
- 64. FSK Modulator (VCO) k1 = Hz/v FSK output NRZ binary input –Δƒ +Δƒ ƒc ƒs ƒm Logic 1 Logic 0 FIGURE 6 FSK modulator where h h-factor (unitless) fm mark frequency (hertz) fs space frequency (hertz) fb bit rate (bits per second) Example 3 Using a Bessel table, determine the minimum bandwidth for the same FSK signal described in Exam- ple 1 with a mark frequency of 49 kHz, a space frequency of 51 kHz, and an input bit rate of 2 kbps. Solution The modulation index is found by substituting into Equation 17: or 1 From a Bessel table, three sets of significant sidebands are produced for a modulation index of one. Therefore, the bandwidth can be determined as follows: B 2(3 1000) 6000 Hz The bandwidth determined in Example 3 using the Bessel table is identical to the bandwidth determined in Example 2. 4-2 FSK Transmitter Figure 6 shows a simplified binary FSK modulator, which is very similar to a conventional FM modulator and is very often a voltage-controlled oscillator (VCO). The center fre- quency (fc) is chosen such that it falls halfway between the mark and space frequencies. A logic 1 input shifts theVCO output to the mark frequency, and a logic 0 input shifts theVCO output to the space frequency. Consequently, as the binary input signal changes back and forth between logic 1 and logic 0 conditions, the VCO output shifts or deviates back and forth between the mark and space frequencies. In a binary FSK modulator, f is the peak frequency deviation of the carrier and is equal to the difference between the carrier rest frequency and either the mark or the space frequency (or half the difference between the carrier rest frequency) and either the mark or the space frequency (or half the difference between the mark and space frequencies).AVCO- FSK modulator can be operated in the sweep mode where the peak frequency deviation is 2 kHz 2 kbps h 冟49 kHz 51 kHz冟 2 kbps Digital Modulation 59
- 65. FIGURE 7 Noncoherent FSK demodulator FIGURE 8 Coherent FSK demodulator simply the product of the binary input voltage and the deviation sensitivity of theVCO.With the sweep mode of modulation, the frequency deviation is expressed mathematically as f vm(t)kl (19) where f peak frequency deviation (hertz) vm(t) peak binary modulating-signal voltage (volts) kl deviation sensitivity (hertz per volt). With binary FSK, the amplitude of the input signal can only be one of two values, one for a logic 1 condition and one for a logic 0 condition. Therefore, the peak frequency devi- ation is constant and always at its maximum value. Frequency deviation is simply plus or minus the peak voltage of the binary signal times the deviation sensitivity of theVCO. Since the peak voltage is the same for a logic 1 as it is for a logic 0, the magnitude of the frequency deviation is also the same for a logic 1 as it is for a logic 0. 4-3 FSK Receiver FSK demodulation is quite simple with a circuit such as the one shown in Figure 7. The FSK input signal is simultaneously applied to the inputs of both bandpass filters (BPFs) through a power splitter. The respective filter passes only the mark or only the space fre- quency on to its respective envelope detector. The envelope detectors, in turn, indicate the total power in each passband, and the comparator responds to the largest of the two pow- ers. This type of FSK detection is referred to as noncoherent detection; there is no frequency involved in the demodulation process that is synchronized either in phase, frequency, or both with the incoming FSK signal. Figure 8 shows the block diagram for a coherent FSK receiver. The incoming FSK sig- nal is multiplied by a recovered carrier signal that has the exact same frequency and phase as the transmitter reference. However, the two transmitted frequencies (the mark and space fre- quencies) are not generally continuous; it is not practical to reproduce a local reference that is coherent with both of them. Consequently, coherent FSK detection is seldom used. Digital Modulation 60
- 66. FIGURE 9 PLL-FSK demodulator FIGURE 10 Noncontinuous FSK waveform The most common circuit used for demodulating binary FSK signals is the phase- locked loop (PLL), which is shown in block diagram form in Figure 9. A PLL-FSK de- modulator works similarly to a PLL-FM demodulator. As the input to the PLL shifts be- tween the mark and space frequencies, the dc error voltage at the output of the phase comparator follows the frequency shift. Because there are only two input frequencies (mark and space), there are also only two output error voltages. One represents a logic 1 and the other a logic 0. Therefore, the output is a two-level (binary) representation of the FSK in- put. Generally, the natural frequency of the PLL is made equal to the center frequency of the FSK modulator. As a result, the changes in the dc error voltage follow the changes in the analog input frequency and are symmetrical around 0 V. Binary FSK has a poorer error performance than PSK or QAM and, consequently, is sel- dom used for high-performance digital radio systems. Its use is restricted to low-performance, low-cost, asynchronous data modems that are used for data communications over analog, voice-band telephone lines. 4-4 Continuous-Phase Frequency-Shift Keying Continuous-phase frequency-shift keying (CP-FSK) is binary FSK except the mark and space frequencies are synchronized with the input binary bit rate. Synchronous simply im- plies that there is a precise time relationship between the two; it does not mean they are equal. With CP-FSK, the mark and space frequencies are selected such that they are separated from the center frequency by an exact multiple of one-half the bit rate (fm and fs n[fb /2]), where n any integer). This ensures a smooth phase transition in the analog output signal when it changes from a mark to a space frequency or vice versa. Figure 10 shows a noncontinuous FSK waveform. It can be seen that when the input changes from a logic 1 to a logic 0 and vice versa, there is an abrupt phase discontinuity in the analog signal. When this occurs, the demodulator has trouble following the frequency shift; consequently, an error may occur. Figure 11 shows a continuous phase FSK waveform. Notice that when the output fre- quency changes, it is a smooth, continuous transition. Consequently, there are no phase dis- continuities. CP-FSK has a better bit-error performance than conventional binary FSK for a given signal-to-noise ratio. The disadvantage of CP-FSK is that it requires synchroniza- tion circuits and is, therefore, more expensive to implement. Digital Modulation 61
- 67. FIGURE 11 Continuous-phase MSK waveform Level converter (UP to BP) Bandpass filter Buffer Balanced modulator Binary data in +v –v +v o Modulated PSK output Reference carrier oscillator sin(ωct) sin(ωct) FIGURE 12 BPSK transmitter 5 PHASE-SHIFT KEYING Phase-shift keying (PSK) is another form of angle-modulated, constant-amplitude digital modulation. PSK is an M-ary digital modulation scheme similar to conventional phase modulation except with PSK the input is a binary digital signal and there are a limited num- ber of output phases possible. The input binary information is encoded into groups of bits before modulating the carrier. The number of bits in a group ranges from 1 to 12 or more. The number of output phases is defined by M as described in Equation 6 and determined by the number of bits in the group (n). 5-1 Binary Phase-Shift Keying The simplest form of PSK is binary phase-shift keying (BPSK), where N 1 and M 2. Therefore, with BPSK, two phases (21 2) are possible for the carrier. One phase repre- sents a logic 1, and the other phase represents a logic 0. As the input digital signal changes state (i.e., from a 1 to a 0 or from a 0 to a 1), the phase of the output carrier shifts between two angles that are separated by 180°. Hence, other names for BPSK are phase reversal key- ing (PRK) and biphase modulation. BPSK is a form of square-wave modulation of a continuous wave (CW) signal. 5-1-1 BPSK transmitter. Figure 12 shows a simplified block diagram of a BPSK transmitter. The balanced modulator acts as a phase reversing switch. Depending on the Digital Modulation 62
- 68. FIGURE 13 (a) Balanced ring modulator; (b) logic 1 input; (c) logic O input logic condition of the digital input, the carrier is transferred to the output either in phase or 180° out of phase with the reference carrier oscillator. Figure 13 shows the schematic diagram of a balanced ring modulator. The balanced modulator has two inputs: a carrier that is in phase with the reference oscillator and the bi- nary digital data. For the balanced modulator to operate properly, the digital input voltage must be much greater than the peak carrier voltage. This ensures that the digital input con- trols the on/off state of diodes D1 to D4. If the binary input is a logic 1 (positive voltage), diodes D1 and D2 are forward biased and on, while diodes D3 and D4 are reverse biased and off (Figure 13b). With the polarities shown, the carrier voltage is developed across Digital Modulation 63
- 69. FIGURE 14 BPSK modulator: (a) truth table; (b) phasor diagram; (c) constellation diagram transformer T2 in phase with the carrier voltage across T1. Consequently, the output signal is in phase with the reference oscillator. If the binary input is a logic 0 (negative voltage), diodes D1 and D2 are reverse biased and off, while diodes D3 and D4 are forward biased and on (Figure 13c).As a result, the car- rier voltage is developed across transformer T2 180° out of phase with the carrier voltage across T1. Consequently, the output signal is 180° out of phase with the reference oscillator. Figure 14 shows the truth table, phasor diagram, and constellation diagram for a BPSK mod- ulator. A constellation diagram, which is sometimes called a signal state-space diagram, is similar to a phasor diagram except that the entire phasor is not drawn. In a constellation di- agram, only the relative positions of the peaks of the phasors are shown. 5-1-2 Bandwidth considerations of BPSK. A balanced modulator is a product modulator; the output signal is the product of the two input signals. In a BPSK modulator, the carrier input signal is multiplied by the binary data. If 1 V is assigned to a logic 1 and 1 V is assigned to a logic 0, the input carrier (sin ωct) is multiplied by either a or 1. Consequently, the output signal is either 1 sin ωct or 1 sin ωct; the first represents a sig- nal that is in phase with the reference oscillator, the latter a signal that is 180° out of phase with the reference oscillator. Each time the input logic condition changes, the output phase changes. Consequently, for BPSK, the output rate of change (baud) is equal to the input rate of change (bps), and the widest output bandwidth occurs when the input binary data are an alternating 1/0 sequence. The fundamental frequency ( fa) of an alternative 1/0 bit sequence is equal to one-half of the bit rate ( fb/2). Mathematically, the output of a BPSK modulator is proportional to BPSK output [sin(2πfat)] [sin(2πfct)] (20) Digital Modulation 64
- 70. FIGURE 15 Output phase-versus-time relationship for a BPSK modulator where fa maximum fundamental frequency of binary input (hertz) fc reference carrier frequency (hertz) Solving for the trig identity for the product of two sine functions, Thus, the minimum double-sided Nyquist bandwidth (B) is and because fa fb/2, where fb input bit rate, where B is the minimum double-sided Nyquist bandwidth. Figure 15 shows the output phase-versus-time relationship for a BPSK waveform. As the figure shows, a logic 1 input produces an analog output signal with a 0° phase angle, and a logic 0 input produces an analog output signal with a 180° phase angle. As the binary input shifts between a logic 1 and a logic 0 condition and vice versa, the phase of the BPSK waveform shifts between 0° and 180°, respectively. For simplicity, only one cycle of the analog carrier is shown in each signaling element, although there may be anywhere be- tween a fraction of a cycle to several thousand cycles, depending on the relationship be- tween the input bit rate and the analog carrier frequency. It can also be seen that the time of one BPSK signaling element (ts) is equal to the time of one information bit (tb), which in- dicates that the bit rate equals the baud. Example 4 For a BPSK modulator with a carrier frequency of 70 MHz and an input bit rate of 10 Mbps, deter- mine the maximum and minimum upper and lower side frequencies, draw the output spectrum, de- termine the minimum Nyquist bandwidth, and calculate the baud. B 2fb 2 fb fc fa 1fc fa 2 or fc fa fc fa 2fa 1 2 cos32π1fc fa 2t4 1 2 cos32π1fc fa 2t4 Digital Modulation 65
- 71. BPF LPF Balanced modulator BPSK input ± sin(ωct) sin(ωct) UP Binary data output Coherent carrier recovery Level converter Clock recovery +v o FIGURE 16 Block diagram of a BPSK receiver Solution Substituting into Equation 20 yields output (sin ωat)(sin ωct) [sin 2π(5 MHz)t][sin 2π(70 MHz)t] lower side frequency upper side frequency Minimum lower side frequency (LSF): LSF 70 MHz 5 MHz 65 MHz Maximum upper side frequency (USF): USF 70 MHz 5 MHz 75 MHz Therefore, the output spectrum for the worst-case binary input conditions is as follows: The minimum Nyquist bandwidth (B) is 1 2 cos 2π 170 MHz 5 MHz2t 1 2 cos 2π 170 MHz 5 MHz2t Digital Modulation 冧 冧 B 75 MHz 65 MHz 10 MHz and the baud fb or 10 megabaud. 5-1-3 BPSK receiver. Figure 16 shows the block diagram of a BPSK receiver. The input signal may be sin ωct or sin ωct. The coherent carrier recovery circuit detects and regenerates a carrier signal that is both frequency and phase coherent with the original transmit carrier. The balanced modulator is a product detector; the output is the product of the two inputs (the BPSK signal and the recovered carrier). The low-pass filter (LPF) sep- arates the recovered binary data from the complex demodulated signal. Mathematically, the demodulation process is as follows. For a BPSK input signal of sin ωct (logic 1), the output of the balanced modulator is output (sin ωct)(sin ωct) sin2 ωct (21) 66
- 72. or leaving It can be seen that the output of the balanced modulator contains a positive voltage ([1/2]V) and a cosine wave at twice the carrier frequency (2 ωc) The LPF has a cutoff fre- quency much lower than 2ωc and, thus, blocks the second harmonic of the carrier and passes only the positive constant component. A positive voltage represents a demodulated logic 1. For a BPSK input signal of sin ωct (logic 0), the output of the balanced modulator is output (sin ωct)(sin ωct) sin2 ωct (filtered out) or leaving The output of the balanced modulator contains a negative voltage ([1/2]V) and a cosine wave at twice the carrier frequency (2ωc). Again, the LPF blocks the second har- monic of the carrier and passes only the negative constant component. A negative voltage represents a demodulated logic 0. 5-2 Quaternary Phase-Shift Keying Quaternary phase shift keying (QPSK), or quadrature PSK as it is sometimes called, is an- other form of angle-modulated, constant-amplitude digital modulation. QPSK is an M-ary encoding scheme where N 2 and M 4 (hence, the name “quaternary” meaning “4”). With QPSK, four output phases are possible for a single carrier frequency. Because there are four output phases, there must be four different input conditions. Because the digital in- put to a QPSK modulator is a binary (base 2) signal, to produce four different input com- binations, the modulator requires more than a single input bit to determine the output con- dition. With two bits, there are four possible conditions: 00, 01, 10, and 11. Therefore, with QPSK, the binary input data are combined into groups of two bits, called dibits. In the mod- ulator, each dibit code generates one of the four possible output phases (45°, 135°, 45°, and 135°). Therefore, for each two-bit dibit clocked into the modulator, a single output change occurs, and the rate of change at the output (baud) is equal to one-half the input bit rate (i.e., two input bits produce one output phase change). 5-2-1 QPSK transmitter. A block diagram of a QPSK modulator is shown in Figure 17. Two bits (a dibit) are clocked into the bit splitter. After both bits have been seri- ally inputted, they are simultaneously parallel outputted. One bit is directed to the I chan- nel and the other to the Q channel. The I bit modulates a carrier that is in phase with the ref- erence oscillator (hence the name “I” for “in phase” channel), and the Q bit modulates a carrier that is 90° out of phase or in quadrature with the reference carrier (hence the name “Q” for “quadrature” channel). It can be seen that once a dibit has been split into the I and Q channels, the operation is the same as in a BPSK modulator. Essentially, a QPSK modulator is two BPSK modula- tors combined in parallel. Again, for a logic 1 1 V and a logic 0 1 V, two phases are possible at the output of the I balanced modulator (sin ωct and sin ωct), and two output 1 2 V logic 0 sin2 ct 1 2 11 cos 2ct2 1 2 1 2 cos 2ct output 1 2 V logic 1 sin2 ct 1 2 11 cos 2ct2 1 2 1 2 cos 2 ct Digital Modulation (filtered out) 67
- 73. FIGURE 17 QPSK modulator phases are possible at the output of the Q balanced modulator (cos ωct and cos ωct). When the linear summer combines the two quadrature (90° out of phase) signals, there are four possible resultant phasors given by these expressions: sin ωct cos ωct, sin ωct cos ωct, sin ωct cos ωct, and sin ωct cos ωct. Example 5 For the QPSK modulator shown in Figure 17, construct the truth table, phasor diagram, and constel- lation diagram. Solution For a binary data input of Q 0 and I 0, the two inputs to the I balanced modulator are 1 and sin ωct, and the two inputs to the Q balanced modulator are 1 and cos ωct. Consequently, the outputs are I balanced modulator (1)(sin ωct) 1 sin ωct Q balanced modulator (1)(cos ωct) 1 cos ωct and the output of the linear summer is 1 cos ωct 1 sin ωct 1.414 sin(ωct 135°) For the remaining dibit codes (01, 10, and 11), the procedure is the same. The results are shown in Figure 18a. In Figures 18b and c, it can be seen that with QPSK each of the four possible output phasors has exactly the same amplitude. Therefore, the binary information must be encoded entirely in the phase of the output signal. This constant amplitude characteristic is the most important characteristic of PSK that distinguishes it from QAM, which is explained later in this chapter. Also, from Figure 18b, it can be seen that the angular separation between any two adjacent phasors in QPSK is 90°. Therefore, a QPSK signal can undergo almost a 45° or 45° shift in phase during transmission and still retain the correct encoded information when demodulated at the receiver. Figure 19 shows the output phase-versus-time relation- ship for a QPSK modulator. Digital Modulation 68
- 74. FIGURE 18 QPSK modulator: (a) truth table; (b) phasor diagram; (c) constellation diagram FIGURE 19 Output phase-versus-time relationship for a QPSK modulator 5-2-2 Bandwidth considerations of QPSK. With QPSK, because the input data are divided into two channels, the bit rate in either the I or the Q channel is equal to one-half of the input data rate (fb/2). (Essentially, the bit splitter stretches the I and Q bits to twice their input bit length.) Consequently, the highest fundamental frequency present at the data input to the I or the Q balanced modulator is equal to one-fourth of the input data rate (one-half of fb/2 fb/4). As a result, the output of the I and Q balanced modulators requires a minimum double-sided Nyquist bandwidth equal to one-half of the incoming bit rate (fN twice fb/4 fb/2). Thus, with QPSK, a bandwidth compression is realized (the minimum bandwidth is less than the incoming bit rate).Also, because the QPSK output signal does not change phase until two bits (a dibit) have been clocked into the bit splitter, the fastest output rate of change (baud) is also equal to one-half of the input bit rate. As with BPSK, the minimum bandwidth and the baud are equal. This relationship is shown in Figure 20. Digital Modulation 69
- 75. FIGURE 20 Bandwidth considerations of a QPSK modulator In Figure 20, it can be seen that the worse-case input condition to the I or Q balanced modulator is an alternative 1/0 pattern, which occurs when the binary input data have a 1100 repetitive pattern. One cycle of the fastest binary transition (a 1/0 sequence) in the I or Q channel takes the same time as four input data bits. Consequently, the highest funda- mental frequency at the input and fastest rate of change at the output of the balanced mod- ulators is equal to one-fourth of the binary input bit rate. The output of the balanced modulators can be expressed mathematically as output (sin ωat)(sin ωct) (22) where Thus, The output frequency spectrum extends from fc fb /4 to fc fb /4, and the minimum band- width (fN) is Example 6 For a QPSK modulator with an input data rate (fb) equal to 10 Mbps and a carrier frequency of 70 MHz, determine the minimum double-sided Nyquist bandwidth (fN) and the baud. Also, compare the results with those achieved with the BPSK modulator in Example 4. Use the QPSK block diagram shown in Figure 17 as the modulator model. Solution The bit rate in both the I and Q channels is equal to one-half of the transmission bit rate, or fbQ fbI fb 2 10 Mbps 2 5 Mbps ¢fc fb 4 ≤ ¢fc fb 4 ≤ 2fb 4 fb 2 1 2 cos 2π¢fc fb 4 ≤t 1 2 cos 2π¢fc fb 4 ≤t output ¢sin 2π fb 4 t≤1sin 2πfct2 at 2π fb 4 t and ct 2πfc Digital Modulation 冧 冧 modulating carrier signal 70
- 76. The highest fundamental frequency presented to either balanced modulator is The output wave from each balanced modulator is (sin 2πfat)(sin 2πfct) The minimum Nyquist bandwidth is B (72.5 67.5) MHz 5 MHz The symbol rate equals the bandwidth; thus, symbol rate 5 megabaud The output spectrum is as follows: B 5 MHz It can be seen that for the same input bit rate the minimum bandwidth required to pass the output of the QPSK modulator is equal to one-half of that required for the BPSK modulator in Example 4.Also, the baud rate for the QPSK modulator is one-half that of the BPSK modulator. The minimum bandwidth for the QPSK system described in Example 6 can also be determined by simply substituting into Equation 10: 5 MHz 5-2-3 QPSK receiver. The block diagram of a QPSK receiver is shown in Figure 21. The power splitter directs the input QPSK signal to the I and Q product detectors and the carrier recovery circuit. The carrier recovery circuit reproduces the original transmit carrier oscillator signal. The recovered carrier must be frequency and phase coherent with the transmit reference carrier. The QPSK signal is demodulated in the I and Q product de- tectors, which generate the original I and Q data bits. The outputs of the product detectors are fed to the bit combining circuit, where they are converted from parallel I and Q data channels to a single binary output data stream. The incoming QPSK signal may be any one of the four possible output phases shown in Figure 18. To illustrate the demodulation process, let the incoming QPSK signal be sin ωct cos ωct. Mathematically, the demodulation process is as follows. B 10 Mbps 2 1 2 cos 2π167.5 MHz2t 1 2 cos 2π172.5 MHz2t 1 2 cos 2π3170 2.52 MHz4t 1 2 cos 2π3170 2.52 MHz4t 1 2 cos 2π1fc fa 2t 1 2 cos 2π1fc fa 2t fa fbQ 2 or fbI 2 5 Mbps 2 2.5 MHz Digital Modulation 71
- 78. The receive QPSK signal (sin ωct cos ωct) is one of the inputs to the I product detector. The other input is the recovered carrier (sin ωct). The output of the I product de- tector is (23) QPSK input signal carrier (sin ct)(sin ct) (cos ct)(sin ct) sin2 ct (cos ct)(sin ct) (filtered out) (equals 0) Again, the receive QPSK signal (sin ωct cos ωct) is one of the inputs to the Q product detector. The other input is the recovered carrier shifted 90° in phase (cos ωct). The output of the Q product detector is (24) QPSK input signal carrier (filtered out) (equals 0) The demodulated I and Q bits (0 and 1, respectively) correspond to the constellation diagram and truth table for the QPSK modulator shown in Figure 18. 5-2-4 Offset QPSK. Offset QPSK (OQPSK) is a modified form of QPSK where the bit waveforms on the I and Q channels are offset or shifted in phase from each other by one- half of a bit time. Figure 22 shows a simplified block diagram, the bit sequence alignment, and the con- stellation diagram for a OQPSK modulator. Because changes in the I channel occur at the midpoints of the Q channel bits and vice versa, there is never more than a single bit change in the dibit code and, therefore, there is never more than a 90° shift in the output phase. In conventional QPSK, a change in the input dibit from 00 to 11 or 01 to 10 causes a corre- sponding 180° shift in the output phase. Therefore, an advantage of OQPSK is the lim- ited phase shift that must be imparted during modulation. A disadvantage of OQPSK is 1 2 V1logic 12 Q 1 2 1 2 cos 2ct 1 2 sin 2ct 1 2 sin 0 1 2 11 cos 2ct2 1 2 sin1c c 2t 1 2 sin1c c 2t cos2 ct 1sin ct2 1cos ct2 Q 1sin ct cos ct2 1cos ct2 1 2 V 1logic 02 I 1 2 1 2 cos 2ct 1 2 sin 2ct 1 2 sin 0 1 2 11 cos 2ct2 1 2 sin1c c 2t 1 2 sin1c c 2t I 1sin ct cos ct2 1sin ct2 Digital Modulation 冦 冦 冦 冦 73
- 79. FIGURE 22 Offset keyed (OQPSK): (a) block diagram; (b) bit alignment; (c) constellation diagram that changes in the output phase occur at twice the data rate in either the I or Q channels. Consequently, with OQPSK the baud and minimum bandwidth are twice that of conven- tional QPSK for a given transmission bit rate. OQPSK is sometimes called OKQPSK (offset-keyed QPSK). 5-3 8-PSK With 8-PSK, three bits are encoded, forming tribits and producing eight different output phases.With 8-PSK, n 3, M 8, and there are eight possible output phases. To encode eight different phases, the incoming bits are encoded in groups of three, called tribits (23 8). 5-3-1 8-PSK transmitter. A block diagram of an 8-PSK modulator is shown in Figure 23. The incoming serial bit stream enters the bit splitter, where it is converted to a parallel, three-channel output (the I or in-phase channel, the Q or in-quadrature chan- nel, and the C or control channel). Consequently, the bit rate in each of the three chan- nels is fb/3. The bits in the I and C channels enter the I channel 2-to-4-level converter, and the bits in the Q and channels enter the Q channel 2-to-4-level converter. Essen- tially, the 2-to-4-level converters are parallel-input digital-to-analog converters (DACs). With two input bits, four output voltages are possible. The algorithm for the DACs is quite simple. The I or Q bit determines the polarity of the output analog sig- nal (logic 1 V and logic 0 V), whereas the C or bit determines the magni- C C Digital Modulation 74
- 80. FIGURE 23 8-PSK modulator FIGURE 24 I- and Q-channel 2-to-4-level converters: (a) I-channel truth table; (b) Q-channel truth table; (c) PAM levels tude (logic 1 1.307 V and logic 0 0.541 V). Consequently, with two magnitudes and two polarities, four different output conditions are possible. Figure 24 shows the truth table and corresponding output conditions for the 2-to-4- level converters. Because the C and bits can never be the same logic state, the outputs from the I and Q 2-to-4-level converters can never have the same magnitude, although they can have the same polarity. The output of a 2-to-4-level converter is an M-ary, pulse- amplitude-modulated (PAM) signal where M 4. Example 7 For a tribit input of Q 0, 1 0, and C 0 (000), determine the output phase for the 8-PSK mod- ulator shown in Figure 23. Solution The inputs to the I channel 2-to-4-level converter are I 0 and C 0. From Figure 24 the output is 0.541 V. The inputs to the Q channel 2-to-4-level converter are Q 0 and Again from Figure 24, the output is 1.307 V. Thus, the two inputs to the I channel product modulators are 0.541 and sin ωct. The output is I (0.541)(sin ωct) 0.541 sin ωct The two inputs to the Q channel product modulator are 1.307 V and cos ωct. The output is Q (1.307)(cos ωct) 1.307 cos ωct C 1. C Digital Modulation 75
- 81. FIGURE 25 8-PSK modulator: (a) truth table; (b) phasor diagram; (c) constellation diagram The outputs of the I and Q channel product modulators are combined in the linear summer and pro- duce a modulated output of summer output 0.541 sin ωct 1.307 cos ωct 1.41 sin(ωct 112.5°) For the remaining tribit codes (001, 010, 011, 100, 101, 110, and 111), the procedure is the same. The results are shown in Figure 25. Digital Modulation 76
- 82. FIGURE 26 Output phase-versus-time relationship for an 8-PSK modulator From Figure 25, it can be seen that the angular separation between any two adjacent phasors is 45°, half what it is with QPSK. Therefore, an 8-PSK signal can undergo almost a 22.5° phase shift during transmission and still retain its integrity. Also, each phasor is of equal magnitude; the tribit condition (actual information) is again contained only in the phase of the signal. The PAM levels of 1.307 and 0.541 are relative values. Any levels may be used as long as their ratio is 0.541/1.307 and their arc tangent is equal to 22.5°. For ex- ample, if their values were doubled to 2.614 and 1.082, the resulting phase angles would not change, although the magnitude of the phasor would increase proportionally. It should also be noted that the tribit code between any two adjacent phases changes by only one bit. This type of code is called the Gray code or, sometimes, the maximum dis- tance code. This code is used to reduce the number of transmission errors. If a signal were to undergo a phase shift during transmission, it would most likely be shifted to an adjacent phasor. Using the Gray code results in only a single bit being received in error. Figure 26 shows the output phase-versus-time relationship of an 8-PSK modulator. 5-3-2 Bandwidth considerations of 8-PSK. With 8-PSK, because the data are di- vided into three channels, the bit rate in the I, Q, or C channel is equal to one-third of the binary input data rate (fb/3). (The bit splitter stretches the I, Q, and C bits to three times their input bit length.) Because the I, Q, and C bits are outputted simultaneously and in parallel, the 2-to-4-level converters also see a change in their inputs (and consequently their outputs) at a rate equal to fb/3. Figure 27 shows the bit timing relationship between the binary input data; the I, Q, and C channel data; and the I and Q PAM signals. It can be seen that the highest fundamental fre- quency in the I, Q, or C channel is equal to one-sixth the bit rate of the binary input (one cy- cle in the I, Q, or C channel takes the same amount of time as six input bits).Also, the highest fundamental frequency in either PAM signal is equal to one-sixth of the binary input bit rate. With an 8-PSK modulator, there is one change in phase at the output for every three data input bits. Consequently, the baud for 8 PSK equals fb /3, the same as the minimum band- width.Again, the balanced modulators are product modulators; their outputs are the product of the carrier and the PAM signal. Mathematically, the output of the balanced modulators is θ (X sin ωat)(sin ωct) (25) where and X 1.307 or 0.541 Thus, X 2 cos 2π¢fc fb 6 ≤t X 2 cos 2π¢fc fb 6 ≤t θ ¢X sin 2π fb 6 t≤1sin 2πfct2 at 2π fb 6 t and ct 2πfct Digital Modulation 冧 冧 modulating signal carrier 77
- 83. FIGURE 27 Bandwidth considerations of an 8-PSK modulator The output frequency spectrum extends from fc fb /6 to fc fb /6, and the minimum band- width (fN) is Example 8 For an 8-PSK modulator with an input data rate (fb) equal to 10 Mbps and a carrier frequency of 70 MHz, determine the minimum double-sided Nyquist bandwidth (fN) and the baud. Also, compare the results with those achieved with the BPSK and QPSK modulators in Examples 4 and 6. Use the 8-PSK block diagram shown in Figure 23 as the modulator model. Solution The bit rate in the I, Q, and C channels is equal to one-third of the input bit rate, or fbC fbQ fbI 10 Mbps 3 3.33 Mbps ¢fc fb 6 ≤ ¢fc fb 6 ≤ 2fb 6 fb 3 Digital Modulation 78
- 84. Therefore, the fastest rate of change and highest fundamental frequency presented to either balanced modulator is The output wave from the balance modulators is (sin 2π fa t)(sin 2πfct) The minimum Nyquist bandwidth is B (71.667 68.333) MHz 3.333 MHz The minimum bandwidth for the 8-PSK can also be determined by simply substituting into Equation 10: 3.33 MHz Again, the baud equals the bandwidth; thus, baud 3.333 megabaud The output spectrum is as follows: B 3.333 MHz It can be seen that for the same input bit rate the minimum bandwidth required to pass the output of an 8-PSK modulator is equal to one-third that of the BPSK modulator in Example 4 and 50% less than that required for the QPSK modulator in Example 6. Also, in each case the baud has been reduced by the same proportions. 5-3-3 8-PSK receiver. Figure 28 shows a block diagram of an 8-PSK receiver. The power splitter directs the input 8-PSK signal to the I and Q product detectors and the car- rier recovery circuit. The carrier recovery circuit reproduces the original reference oscilla- tor signal. The incoming 8-PSK signal is mixed with the recovered carrier in the I product detector and with a quadrature carrier in the Q product detector. The outputs of the product detectors are 4-level PAM signals that are fed to the 4-to-2-level analog-to-digital con- verters (ADCs). The outputs from the I channel 4-to-2-level converter are the I and C bits, whereas the outputs from the Q channel 4-to-2-level converter are the Q and bits. The parallel-to-serial logic circuit converts the I/C and bit pairs to serial I, Q, and C output data streams. 5-4 16-PSK 16-PSK is an M-ary encoding technique where M 16; there are 16 different output phases possible. With 16-PSK, four bits (called quadbits) are combined, producing 16 different output phases. With 16-PSK, n 4 and M 16; therefore, the minimum bandwidth and QC C B 10 Mbps 3 1 2 cos2π168.333 MHz2t 1 2 cos 2π171.667 MHz2t 1 2 cos 2π3170 1.6672 MHz4t 1 2 cos 2π3170 1.6672 MHz4t 1 2 cos 2π1fc fa 2t 1 2 cos 2π1fc fa 2t fa fbC 2 or fbQ 2 or fbI 2 3.33 Mbps 2 1.667 Mbps Digital Modulation 79
- 86. FIGURE 29 16-PSK: (a) truth table; (b) constellation diagram baud equal one-fourth the bit rate ( fb/4). Figure 29 shows the truth table and constellation diagram for 16-PSK, respectively. Comparing Figures 18, 25, and 29 shows that as the level of encoding increases (i.e., the values of n and M increase), more output phases are possi- ble and the closer each point on the constellation diagram is to an adjacent point. With 16- PSK, the angular separation between adjacent output phases is only 22.5°. Therefore, 16- PSK can undergo only a 11.25° phase shift during transmission and still retain its integrity. For an M-ary PSK system with 64 output phases (n 6), the angular separation between adjacent phases is only 5.6°. This is an obvious limitation in the level of encoding (and bit rates) possible with PSK, as a point is eventually reached where receivers cannot discern the phase of the received signaling element. In addition, phase impairments inherent on communications lines have a tendency to shift the phase of the PSK signal, destroying its integrity and producing errors. 6 QUADRATURE-AMPLITUDE MODULATION Quadrature-amplitude modulation (QAM) is a form of digital modulation similar to PSK except the digital information is contained in both the amplitude and the phase of the trans- mitted carrier. With QAM, amplitude and phase-shift keying are combined in such a way that the positions of the signaling elements on the constellation diagrams are optimized to achieve the greatest distance between elements, thus reducing the likelihood of one element being misinterpreted as another element. Obviously, this reduces the likelihood of errors oc- curring. 6-1 8-QAM 8-QAM is an M-ary encoding technique where M 8. Unlike 8-PSK, the output signal from an 8-QAM modulator is not a constant-amplitude signal. 6-1-1 8-QAM transmitter. Figure 30a shows the block diagram of an 8-QAM transmitter. As you can see, the only difference between the 8-QAM transmitter and the 8- PSK transmitter shown in Figure 23 is the omission of the inverter between the C channel and the Q product modulator. As with 8-PSK, the incoming data are divided into groups of three bits (tribits): the I, Q, and C bit streams, each with a bit rate equal to one-third of Digital Modulation 81
- 87. the incoming data rate. Again, the I and Q bits determine the polarity of the PAM signal at the output of the 2-to-4-level converters, and the C channel determines the magnitude. Be- cause the C bit is fed uninverted to both the I and the Q channel 2-to-4-level converters, the magnitudes of the I and Q PAM signals are always equal. Their polarities depend on the logic condition of the I and Q bits and, therefore, may be different. Figure 30b shows the truth table for the I and Q channel 2-to-4-level converters; they are identical. Example 9 For a tribit input of Q 0, I 0, and C 0 (000), determine the output amplitude and phase for the 8-QAM transmitter shown in Figure 30a. Solution The inputs to the I channel 2-to-4-level converter are I 0 and C 0. From Figure 30b, the output is 0.541 V. The inputs to the Q channel 2-to-4-level converter are Q 0 and C 0. Again from Figure 30b, the output is 0.541 V. Thus, the two inputs to the I channel product modulator are 0.541 and sin ωct. The output is I (0.541)(sin ωct) 0.541 sin ωct The two inputs to the Q channel product modulator are 0.541 and cos ωct. The output is Q (0.541)(cos ωct) 0.541 cos ωct The outputs from the I and Q channel product modulators are combined in the linear summer and pro- duce a modulated output of summer output 0.541 sin ωct 0.541 cos ωct 0.765 sin(ωct 135°) For the remaining tribit codes (001, 010, 011, 100, 101, 110, and 111), the procedure is the same. The results are shown in Figure 31. Figure 32 shows the output phase-versus-time relationship for an 8-QAM modulator. Note that there are two output amplitudes, and only four phases are possible. 6-1-2 Bandwidth considerations of 8-QAM. In 8-QAM, the bit rate in the I and Q channels is one-third of the input binary rate, the same as in 8-PSK. As a result, the high- est fundamental modulating frequency and fastest output rate of change in 8-QAM are the same as with 8-PSK. Therefore, the minimum bandwidth required for 8-QAM is fb/3, the same as in 8-PSK. Digital Modulation FIGURE 30 8-QAM transmitter: (a) block diagram; (b) truth table 4 level converters 82
- 88. FIGURE 31 8-QAM modulator: (a) truth table; (b) phasor diagram; (c) constellation diagram FIGURE 32 Output phase and amplitude-versus-time relationship for 8-QAM 6-1-3 8-QAM receiver. An 8-QAM receiver is almost identical to the 8-PSK re- ceiver shown in Figure 28. The differences are the PAM levels at the output of the product detectors and the binary signals at the output of the analog-to-digital converters. Because there are two transmit amplitudes possible with 8-QAM that are different from those achievable with 8-PSK, the four demodulated PAM levels in 8-QAM are different from those in 8-PSK. Therefore, the conversion factor for the analog-to-digital converters must also be different. Also, with 8-QAM the binary output signals from the I channel analog- to-digital converter are the I and C bits, and the binary output signals from the Q channel analog-to-digital converter are the Q and C bits. Digital Modulation 83
- 89. FIGURE 33 16-QAM transmitter block diagram FIGURE 34 Truth tables for the I- and Q-channel 2-to-4- level converters: (a) I channel; (b) Q channel 6-2 16-QAM As with the 16-PSK, 16-QAM is an M-ary system where M 16. The input data are acted on in groups of four (24 16). As with 8-QAM, both the phase and the amplitude of the transmit carrier are varied. 6-2-1 QAM transmitter. The block diagram for a 16-QAM transmitter is shown in Figure 33. The input binary data are divided into four channels: I, I , Q, and Q . The bit rate in each channel is equal to one-fourth of the input bit rate ( fb/4). Four bits are serially clocked into the bit splitter; then they are outputted simultaneously and in parallel with the I, I , Q, and Q channels. The I and Q bits determine the polarity at the output of the 2-to- 4-level converters (a logic 1 positive and a logic 0 negative). The I and Q bits deter- mine the magnitude (a logic I 0.821 V and a logic 0 0.22 V). Consequently, the 2-to- 4-level converters generate a 4-level PAM signal. Two polarities and two magnitudes are possible at the output of each 2-to-4-level converter. They are 0.22 V and 0.821 V. The PAM signals modulate the in-phase and quadrature carriers in the product mod- ulators. Four outputs are possible for each product modulator. For the I product modulator, they are 0.821 sin ωct, 0.821 sin ωct, 0.22 sin ωct, and 0.22 sin ωct. For the Q prod- uct modulator, they are 0.821 cos ωct, 0.22 cos ωct, 0.821 cos ωct, and 0.22 cos ωct. The linear summer combines the outputs from the I and Q channel product modulators and produces the 16 output conditions necessary for 16-QAM. Figure 34 shows the truth table for the I and Q channel 2-to-4-level converters. Digital Modulation 84
- 90. FIGURE 35 16-QAM modulator: (a) truth table; (b) phasor diagram; (c) constellation diagram Example 10 For a quadbit input of I 0, I 0, Q 0, and Q 0 (0000), determine the output amplitude and phase for the 16-QAM modulator shown in Figure 33. Solution The inputs to the I channel 2-to-4-level converter are I 0 and I 0. From Figure 34, the output is 0.22 V. The inputs to the Q channel 2-to-4-level converter are Q 0 and Q 0.Again from Figure 34, the output is 0.22 V. Thus, the two inputs to the I channel product modulator are 0.22 V and sin ωct. The output is I (0.22)(sin ωct) 0.22 sin ωct The two inputs to the Q channel product modulator are 0.22 V and cos ωct. The output is Q (0.22)(cos ωct) 0.22 cos ωct The outputs from the I and Q channel product modulators are combined in the linear summer and pro- duce a modulated output of summer output 0.22 sin ωct 0.22 cos ωct 0.311 sin(ωct 135°) For the remaining quadbit codes, the procedure is the same. The results are shown in Figure 35. Digital Modulation 85
- 91. FIGURE 36 Bandwidth considerations of a 16-QAM modulator 6-2-2 Bandwidth considerations of 16-QAM. With a 16-QAM, because the input data are divided into four channels, the bit rate in the I, I , Q, or Q channel is equal to one- fourth of the binary input data rate ( fb/4). (The bit splitter stretches the I, I , Q, and Q bits to four times their input bit length.) Also, because the I, I , Q, and Q bits are outputted si- multaneously and in parallel, the 2-to-4-level converters see a change in their inputs and outputs at a rate equal to one-fourth of the input data rate. Figure 36 shows the bit timing relationship between the binary input data; the I, I , Q, and Q channel data; and the I PAM signal. It can be seen that the highest fundamental frequency in the I, I , Q, or Q channel is equal to one-eighth of the bit rate of the binary in- put data (one cycle in the I, I , Q, or Q channel takes the same amount of time as eight in- put bits). Also, the highest fundamental frequency of either PAM signal is equal to one- eighth of the binary input bit rate. With a 16-QAM modulator, there is one change in the output signal (either its phase, amplitude, or both) for every four input data bits. Consequently, the baud equals fb/4, the same as the minimum bandwidth. Digital Modulation 86
- 92. Digital Modulation Again, the balanced modulators are product modulators and their outputs can be rep- resented mathematically as output (X sin ωat)(sin ωct) (26) where and X 0.22 or 0.821 Thus, The output frequency spectrum extends from fc fb/8 to fc fb/8, and the minimum band- width (fN) is Example 11 For a 16-QAM modulator with an input data rate (fb) equal to 10 Mbps and a carrier frequency of 70 MHz, determine the minimum double-sided Nyquist frequency (fN) and the baud. Also, compare the results with those achieved with the BPSK, QPSK, and 8-PSK modulators in Examples 4, 6, and 8. Use the 16-QAM block diagram shown in Figure 33 as the modulator model. Solution The bit rate in the I, I , Q, and Q channels is equal to one-fourth of the input bit rate, or Therefore, the fastest rate of change and highest fundamental frequency presented to either balanced modulator is The output wave from the balanced modulator is (sin 2π fat)(sin 2π fct) The minimum Nyquist bandwidth is B (71.25 68.75) MHz 2.5 MHz The minimum bandwidth for the 16-QAM can also be determined by simply substituting into Equation 10: 2.5 MHz B 10 Mbps 4 1 2 cos 2π168.75 MHz2t 1 2 cos 2π171.25 MHz2t 1 2 cos 2π3170 1.252 MHz4t 1 2 cos 2π3170 1.252 MHz4t 1 2 cos 2π1fc fa 2t 1 2 cos 2π1fc fa 2t fa fbI 2 or fbI¿ 2 or fbQ 2 or fbQ¿ 2 2.5 Mbps 2 1.25 MHz fbI fbI¿ fbQ fbQ¿ fb 4 10 Mbps 4 2.5 Mbps ¢fc fb 8 ≤ ¢fc fb 8 ≤ 2fb 8 fb 4 X 2 cos 2π ¢fc fb 8 ≤t X 2 cos 2π ¢fc fb 8 ≤t output ¢X sin 2π fb 8 t≤1sin 2πfct2 at 2π fb 8 t and c t 2πfct 冧 冧 modulating signal carrier 87
- 93. Digital Modulation Table 1 ASK, FSK, PSK, and QAM Summary Modulation Encoding Scheme Outputs Possible Minimum Bandwidth Baud ASK Single bit 2 fb fb FSK Single bit 2 fb fb BPSK Single bit 2 fb fb QPSK Dibits 4 fb /2 fb /2 8-PSK Tribits 8 fb /3 fb /3 8-QAM Tribits 8 fb /3 fb /3 16-QAM Quadbits 16 fb /4 fb /4 16-PSK Quadbits 16 fb /4 fb /4 32-PSK Five bits 32 fb /5 fb /5 32-QAM Five bits 32 fb /5 fb /5 64-PSK Six bits 64 fb /6 fb /6 64-QAM Six bits 64 fb /6 fb /6 128-PSK Seven bits 128 fb /7 fb /7 128-QAM Seven bits 128 fb /7 fb /7 Note: fb indicates a magnitude equal to the input bit rate. The symbol rate equals the bandwidth; thus, symbol rate 2.5 megabaud The output spectrum is as follows: B 2.5 MHz For the same input bit rate, the minimum bandwidth required to pass the output of a 16-QAM mod- ulator is equal to one-fourth that of the BPSK modulator, one-half that of QPSK, and 25% less than with 8-PSK. For each modulation technique, the baud is also reduced by the same proportions. Example 12 For the following modulation schemes, construct a table showing the number of bits encoded, num- ber of output conditions, minimum bandwidth, and baud for an information data rate of 12 kbps: QPSK, 8-PSK, 8-QAM, 16-PSK, and 16-QAM. Solution Modulation n M B (Hz) baud QPSK 2 4 6000 6000 8-PSK 3 8 4000 4000 8-QAM 3 8 4000 4000 16-PSK 4 16 3000 3000 16-QAM 4 16 3000 3000 From Example 12, it can be seen that a 12-kbps data stream can be propagated through a narrower bandwidth using either 16-PSK or 16-QAM than with the lower levels of encoding. Table 1 summarizes the relationship between the number of bits encoded, the num- ber of output conditions possible, the minimum bandwidth, and the baud for ASK, FSK, PSK, and QAM. Note that with the three binary modulation schemes (ASK, FSK, and 88
- 94. Digital Modulation BPSK), n 1, M 2, only two output conditions are possible, and the baud is equal to the bit rate. However, for values of n 1, the number of output conditions increases, and the minimum bandwidth and baud decrease. Therefore, digital modulation schemes where n 1 achieve bandwidth compression (i.e., less bandwidth is required to propagate a given bit rate). When data compression is performed, higher data transmission rates are possible for a given bandwidth. 7 BANDWIDTH EFFICIENCY Bandwidth efficiency (sometimes called information density or spectral efficiency) is often used to compare the performance of one digital modulation technique to another. In essence, bandwidth efficiency is the ratio of the transmission bit rate to the minimum band- width required for a particular modulation scheme. Bandwidth efficiency is generally nor- malized to a 1-Hz bandwidth and, thus, indicates the number of bits that can be propagated through a transmission medium for each hertz of bandwidth. Mathematically, bandwidth efficiency is Bη transmission bit rate (bps) (27) minimum bandwidth (Hz) bits/s bits/s bits hertz cycles/s cycle where Bη bandwidth efficiency Bandwidth efficiency can also be given as a percentage by simply multiplying Bη by 100. Example 13 For an 8-PSK system, operating with an information bit rate of 24 kbps, determine (a) baud, (b) min- imum bandwidth, and (c) bandwidth efficiency. Solution a. Baud is determined by substituting into Equation 10: b. Bandwidth is determined by substituting into Equation 11: c. Bandwidth efficiency is calculated from Equation 27: 3 bits per second per cycle of bandwidth Example 14 For 16-PSK and a transmission system with a 10 kHz bandwidth, determine the maximum bit rate. Solution The bandwidth efficiency for 16-PSK is 4, which means that four bits can be propagated through the system for each hertz of bandwidth. Therefore, the maximum bit rate is simply the prod- uct of the bandwidth and the bandwidth efficiency, or bit rate 4 10,000 40,000 bps Bh 24,000 8000 bps Hz B 24,000 3 8000 baud 24,000 3 8000 89
- 95. Digital Modulation FIGURE 37 Squaring loop carrier recovery circuit for a BPSK receiver Table 2 ASK, FSK, PSK, and QAM Summary Modulation Encoding Scheme Outputs Possible Minimum Bandwidth Baud Bη ASK Single bit 2 fb fb 1 FSK Single bit 2 fb fb 1 BPSK Single bit 2 fb fb 1 QPSK Dibits 4 fb /2 fb /2 2 8-PSK Tribits 8 fb /3 fb /3 3 8-QAM Tribits 8 fb /3 fb /3 3 16-PSK Quadbits 16 fb /4 fb /4 4 16-QAM Quadbits 16 fb /4 fb /4 4 32-PSK Five bits 32 fb /5 fb /5 5 64-QAM Six bits 64 fb /6 fb /6 6 Note: fb indicates a magnitude equal to the input bit rate. 7-1 Digital Modulation Summary The properties of several digital modulation schemes are summarized in Table 2. 8 CARRIER RECOVERY Carrier recovery is the process of extracting a phase-coherent reference carrier from a re- ceiver signal. This is sometimes called phase referencing. In the phase modulation techniques described thus far, the binary data were encoded as a precise phase of the transmitted carrier. (This is referred to as absolute phase encoding.) Depending on the encoding method, the angular separation between adjacent phasors varied between 30° and 180°. To correctly demodulate the data, a phase-coherent carrier was re- covered and compared with the received carrier in a product detector. To determine the ab- solute phase of the received carrier, it is necessary to produce a carrier at the receiver that is phase coherent with the transmit reference oscillator. This is the function of the carrier re- covery circuit. With PSK and QAM, the carrier is suppressed in the balanced modulators and, there- fore, is not transmitted. Consequently, at the receiver the carrier cannot simply be tracked with a standard phase-locked loop (PLL). With suppressed-carrier systems, such as PSK and QAM, sophisticated methods of carrier recovery are required, such as a squaring loop, a Costas loop, or a remodulator. 8-1 Squaring Loop A common method of achieving carrier recovery for BPSK is the squaring loop. Figure 37 shows the block diagram of a squaring loop. The received BPSK waveform is filtered and then squared. The filtering reduces the spectral width of the received noise. The squaring circuit removes the modulation and generates the second harmonic of the carrier frequency. This harmonic is phase tracked by the PLL. The VCO output frequency from the PLL then is divided by 2 and used as the phase reference for the product detectors. 90
- 96. Digital Modulation FIGURE 38 Costas loop carrier recovery circuit With BPSK, only two output phases are possible: sin ωct and sin ωct. Mathe- matically, the operation of the squaring circuit can be described as follows. For a receive signal of sin ωct, the output of the squaring circuit is output (sin ωct)(sin ωct) sin2 ωct (filtered out) For a received signal of sin ωct, the output of the squaring circuit is output (sin ωct)(sin ωct) sin2 ωct (filtered out) It can be seen that in both cases, the output from the squaring circuit contained a con- stant voltage (1/2 V) and a signal at twice the carrier frequency (cos 2ωct). The constant voltage is removed by filtering, leaving only cos 2ωct. 8-2 Costas Loop A second method of carrier recovery is the Costas, or quadrature, loop shown in Figure 38. The Costas loop produces the same results as a squaring circuit followed by an ordinary PLL in place of the BPF. This recovery scheme uses two parallel tracking loops (I and Q) simul- taneously to derive the product of the I and Q components of the signal that drives the VCO. The in-phase (I) loop uses the VCO as in a PLL, and the quadrature (Q) loop uses a 90° shifted VCO signal. Once the frequency of the VCO is equal to the suppressed-carrier 1 2 11 cos 2ct2 1 2 1 2 cos 2ct 1 2 11 cos 2ct2 1 2 1 2 cos 2ct 91
- 97. Digital Modulation FIGURE 39 Remodulator loop carrier recovery circuit frequency, the product of the I and Q signals will produce an error voltage proportional to any phase error in the VCO. The error voltage controls the phase and, thus, the frequency of the VCO. 8-3 Remodulator A third method of achieving recovery of a phase and frequency coherent carrier is the re- modulator, shown in Figure 39. The remodulator produces a loop error voltage that is pro- portional to twice the phase error between the incoming signal and the VCO signal. The re- modulator has a faster acquisition time than either the squaring or the Costas loops. Carrierrecoverycircuitsforhigher-than-binaryencodingtechniquesaresimilartoBPSK exceptthatcircuitsthatraisethereceivesignaltothefourth,eighth,andhigherpowersareused. 9 CLOCK RECOVERY As with any digital system, digital radio requires precise timing or clock synchronization between the transmit and the receive circuitry. Because of this, it is necessary to regenerate clocks at the receiver that are synchronous with those at the transmitter. Figure 40a shows a simple circuit that is commonly used to recover clocking infor- mation from the received data. The recovered data are delayed by one-half a bit time and then compared with the original data in an XOR circuit. The frequency of the clock that is recovered with this method is equal to the received data rate (fb). Figure 40b shows the re- lationship between the data and the recovered clock timing. From Figure 40b, it can be seen that as long as the receive data contain a substantial number of transitions (1/0 sequences), the recovered clock is maintained. If the receive data were to undergo an extended period of successive 1s or 0s, the recovered clock would be lost. To prevent this from occurring, the data are scrambled at the transmit end and descrambled at the receive end. Scrambling introduces transitions (pulses) into the binary signal using a prescribed algorithm, and the descrambler uses the same algorithm to remove the transitions. 92
- 98. Digital Modulation FIGURE 40 (a) Clock recovery circuit; (b) timing diagram FIGURE 41 DBPSK modulator: (a) block diagram; (b) timing diagram 10 DIFFERENTIAL PHASE-SHIFT KEYING Differential phase-shift keying (DPSK) is an alternative form of digital modulation where the binary input information is contained in the difference between two successive sig- naling elements rather than the absolute phase. With DPSK, it is not necessary to recover a phase-coherent carrier. Instead, a received signaling element is delayed by one signal- ing element time slot and then compared with the next received signaling element. The difference in the phase of the two signaling elements determines the logic condition of the data. 10-1 Differential BPSK 10-1-1 DBPSK transmitter. Figure 41a shows a simplified block diagram of a differential binary phase-shift keying (DBPSK) transmitter. An incoming information bit is 93
- 99. Digital Modulation FIGURE 42 DBPSK demodulator: (a) block diagram; (b) timing sequence XNORed with the preceding bit prior to entering the BPSK modulator (balanced modulator). For the first data bit, there is no preceding bit with which to compare it.Therefore, an initial ref- erencebitisassumed.Figure41bshowstherelationshipbetweentheinputdata,theXNORout- put data, and the phase at the output of the balanced modulator. If the initial reference bit is as- sumed a logic 1, the output from the XNOR circuit is simply the complement of that shown. In Figure 41b, the first data bit is XNORed with the reference bit. If they are the same, the XNOR output is a logic 1; if they are different, the XNOR output is a logic 0. The bal- anced modulator operates the same as a conventional BPSK modulator; a logic 1 produces sin ωct at the output, and a logic 0 produces sin ωct at the output. 10-1-2 DBPSK receiver. Figure 42 shows the block diagram and timing sequence for a DBPSK receiver. The received signal is delayed by one bit time, then compared with the next signaling element in the balanced modulator. If they are the same, a logic 1 ( volt- age) is generated. If they are different, a logic 0 ( voltage) is generated. If the reference phase is incorrectly assumed, only the first demodulated bit is in error. Differential encod- ing can be implemented with higher-than-binary digital modulation schemes, although the differential algorithms are much more complicated than for DBPSK. The primary advantage of DBPSK is the simplicity with which it can be imple- mented. With DBPSK, no carrier recovery circuit is needed. A disadvantage of DBPSK is that it requires between 1 dB and 3 dB more signal-to-noise ratio to achieve the same bit er- ror rate as that of absolute PSK. 11 TRELLIS CODE MODULATION Achieving data transmission rates in excess of 9600 bps over standard telephone lines with approximately a 3-kHz bandwidth obviously requires an encoding scheme well beyond the quadbits used with 16-PSK or 16-QAM (i.e., M must be significantly greater than 16). As might be expected, higher encoding schemes require higher signal-to-noise ratios. Using the Shannon limit for information capacity (Equation 4), a data transmission rate of 28.8 kbps through a 3200-Hz bandwidth requires a signal-to-noise ratio of I(bps) (3.32 B) log(1 S/N) 94
- 100. Digital Modulation therefore, 28.8 kbps (3.32)(3200) log(1 S/N) 28,800 10,624 log(1 S/N) 2.71 log(1 S/N) thus, 102.71 1 S/N 513 1 S/N 512 S/N in dB, S/N(dB) 10 log 512 27 dB Transmission rates of 56 kbps require a signal-to-noise ratio of 53 dB, which is virtually impossible to achieve over a standard telephone circuit. Data transmission rates in excess of 56 kbps can be achieved, however, over standard telephone circuits using an encoding technique called trellis code modulation (TCM). Dr. Ungerboeck at IBM Zuerich Research Laboratory developed TCM, which involves us- ing convolutional (tree) codes, which combines encoding and modulation to reduce the probability of error, thus improving the bit error performance. The fundamental idea behind TCM is introducing controlled redundancy in the bit stream with a convolutional code, which reduces the likelihood of transmission errors. What sets TCM apart from standard encoding schemes is the introduction of redundancy by doubling the number of signal points in a given PSK or QAM constellation. Trelliscodemodulationissometimesthoughtofasamagicalmethodofincreasingtrans- missionbitratesovercommunicationssystemsusingQAMorPSKwithfixedbandwidths.Few people fully understand this concept, as modem manufacturers do not seem willing to share in- formation on TCM. Therefore, the following explanation is intended not to fully describe the process of TCM but rather to introduce the topic and give the reader a basic understanding of how TCM works and the advantage it has over conventional digital modulation techniques. M-ary QAM and PSK utilize a signal set of 2N M, where N equals the number of bits encoded into M different conditions. Therefore, N 2 produces a standard PSK con- stellation with four signal points (i.e., QPSK) as shown in Figure 43a. Using TCM, the number of signal points increases to two times M possible symbols for the same factor-of- M reduction in bandwidth while transmitting each signal during the same time interval. TCM-encoded QPSK is shown in Figure 43b. Trellis coding also defines the manner in which signal-state transitions are allowed to occur, and transitions that do not follow this pattern are interpreted in the receiver as trans- mission errors. Therefore, TCM can improve error performance by restricting the manner in which signals are allowed to transition. For values of N greater than 2, QAM is the mod- ulation scheme of choice for TCM; however, for simplification purposes, the following ex- planation uses PSK as it is easier to illustrate. Figure 44 shows a TCM scheme using two-state 8-PSK, which is essentially two QPSK constellations offset by 45°. One four-state constellation is labeled 0-4-2-6, and the other is labeled 1-5-3-7. For this explanation, the signal point labels 0 through 7 are meant not to represent the actual data conditions but rather to simply indicate a convenient method of labeling the various signal points. Each digit represents one of four signal points per- mitted within each of the two QPSK constellations. When in the 0-4-2-6 constellation and a 0 or 4 is transmitted, the system remains in the same constellation. However, when either a 2 or 6 is transmitted, the system switches to the 1-5-3-7 constellation. Once in the 1-5-3-7 28,800 10,624 log11 SN2 95
- 101. Digital Modulation 0-4-2-6 0 4 0 4 1 2 5 6 3 7 1 2 5 6 3 7 1-5-3-7 FIGURE 44 8-PSK TCM constellations FIGURE 43 QPSK constellations: (a) standard encoding format; (b) trellis encoding format constellation and a 3 or 7 is transmitted, the system remains in the same constellation, and if a 1 or 5 is transmitted, the system switches to the 0-4-2-6 constellation. Remember that each symbol represents two bits, so the system undergoes a 45° phase shift whenever it switches between the two constellations. A complete error analysis of standard QPSK compared with TCM QPSK would reveal a coding gain for TCM of 2-to-1 1 or 3 dB. Table 3 lists the coding gains achieved for TCM coding schemes with several different trellis states. The maximum data rate achievable using a given bandwidth can be determined by re- arranging Equation 10: N B fb 96
- 102. Digital Modulation Table 3 Trellis Coding Gain Number of Trellis States Coding Gain (dB) 2 3.0 4 5.5 8 6.0 16 6.5 32 7.1 64 7.3 128 7.3 256 7.4 where N number of bits encoded (bits) B bandwidth (hertz) fb transmission bit rate (bits per second) Remember that with M-ary QAM or PSK systems, the baud equals the minimum re- quired bandwidth. Therefore, a 3200-Hz bandwidth using a nine-bit trellis code pro- duces a 3200 baud signal with each baud carrying nine bits. Therefore, the transmission rate fb 9 3200 28.8 kbps. TCM is thought of as a coding scheme that improves on standard QAM by increasing the distance between symbols on the constellation (known as the Euclidean distance). The first TCM system used a five-bit code, which included four QAM bits (a quadbit) and a fifth bit used to help decode the quadbit. Transmitting five bits within a single signaling element requires producing 32 discernible signals. Figure 45 shows a 128-point QAM constallation. 90° 270° 0° 180° 8 6 4 2 0110000 1100000 0010011 1011101 0110101 1100101 0010001 1011000 0001000 1100111 0101000 1010000 0000000 1100011 0101101 1010101 0000101 1100001 0101000 0100000 1111000 1101001 1110000 0110001 1001011 1010111 1001111 1010011 1001101 0011011 0101100 1111101 1100010 1111100 1101010 1110100 1101011 1110101 0110011 0000100 0110010 1001001 1010100 1001010 1010100 1001110 1010010 1001100 1010001 1001000 0101000 0101000 0101000 0101000 0101000 0101000 0101000 1111111 1100110 1111110 1101110 1110110 1101111 1110111 0110111 0001100 0110110 1000001 1011100 1000010 1011110 1000110 1011010 1000100 1011001 100000 0101011 0010010 0101010 0100010 0010100 0100011 0010000 1111010 1100100 1111010 11001100 1110010 1101101 1110011 0001101 0110100 0111100 0001011 0111101 1000011 1011111 1000111 1011011 1000101 0101001 0100001 0010101 1111001 1101000 1110001 0111000 –2 –8 –6 –4 –2 2 4 6 8 –4 –6 –8 0010111 0001111 0010110 0001110 0111110 0001010 0111111 0001001 0101001 0000110 0111010 0000010 0111011 0000001 0011111 0100100 0011101 0100101 0000111 0111001 0000011 FIGURE 45 128-Point QAM TCM constellation 97
- 103. Digital Modulation 236 234 226 229 201 222 203 194 238 186 142 103 69 43 22 9 1 0 5 16 32 56 85 122 163 213 190 144 106 73 45 25 11 3 2 7 18 36 59 88 124 166 217 199 152 113 80 52 33 19 12 10 14 26 42 66 97 134 174 225 210 167 128 94 67 47 34 27 23 29 40 57 81 111 147 187 237 232 183 149 115 89 68 53 46 44 51 61 78 99 132 168 209 214 175 139 116 95 82 74 70 76 86 104 129 157 195 205 176 150 130 114 107 105 109 120 136 161 191 215 184 169 153 145 143 151 159 178 233 211 200 192 188 239 196 204 223 202 231 227 235 148 108 75 50 28 13 6 4 8 21 38 63 93 127 171 219 158 119 84 60 39 24 17 15 20 30 49 72 101 138 182 230 177 135 102 77 55 41 35 31 37 48 65 91 118 155 198 160 126 98 79 64 58 54 62 71 90 112 141 180 221 189 156 131 110 96 87 33 92 100 117 140 172 208 193 165 146 133 123 121 125 137 154 179 207 205 185 173 164 162 170 181 197 220 224 216 212 218 228 FIGURE 46 One-fourth of a 960-Point QAM TCM constellation A 3200-baud signal using nine-bit TCM encoding produces 512 different codes. The nine data bits plus a redundant bit for TCM requires a 960-point constellation. Figure 46 shows one-fourth of the 960-point superconstellation showing 240 signal points. The full superconstellation can be obtained by rotating the 240 points shown by 90°, 180°, and 270°. 12 PROBABILITY OF ERROR AND BIT ERROR RATE Probability of error P(e) and bit error rate (BER) are often used interchangeably, al- though in practice they do have slightly different meanings. P(e) is a theoretical (mathe- matical) expectation of the bit error rate for a given system. BER is an empirical (histor- ical) record of a system’s actual bit error performance. For example, if a system has a P(e) of 105 , this means that mathematically you can expect one bit error in every 100,000 bits transmitted (1/105 1/100,000). If a system has a BER of 105 , this means that in past performance there was one bit error for every 100,000 bits transmit- ted. A bit error rate is measured and then compared with the expected probability of er- ror to evaluate a system’s performance. 98
- 104. Digital Modulation Probability of error is a function of the carrier-to-noise power ratio (or, more specif- ically, the average energy per bit-to-noise power density ratio) and the number of possible encoding conditions used (M-ary). Carrier-to-noise power ratio is the ratio of the average carrier power (the combined power of the carrier and its associated sidebands) to the thermal noise power. Carrier power can be stated in watts or dBm, where (28) Thermal noise power is expressed mathematically as N KTB (watts) (29) where N thermal noise power (watts) K Boltzmann’s proportionality constant (1.38 1023 joules per kelvin) T temperature (kelvin: 0 K 273° C, room temperature 290 K) B bandwidth (hertz) Stated in dBm, (30) Mathematically, the carrier-to-noise power ratio is (31) where C carrier power (watts) N noise power (watts) Stated in dB, C(dBm) N(dBm) (32) Energy per bit is simply the energy of a single bit of information. Mathematically, en- ergy per bit is Eb CTb (J/bit) (33) where Eb energy of a single bit (joules per bit) Tb time of a single bit (seconds) C carrier power (watts) Stated in dBJ, Eb(dBJ) 10 log Eb (34) and because Tb 1/fb, where fb is the bit rate in bits per second, Eb can be rewritten as (35) Stated in dBJ, (36) 10 log C 10 log fb (37) Noise power density is the thermal noise power normalized to a 1-Hz bandwidth (i.e., the noise power present in a 1-Hz bandwidth). Mathematically, noise power density is Eb 1dBJ2 10 log C fb Eb C fb 1Jbit2 C N 1dB2 10 log C N C N C KTB 1unitless ratio2 N1dBm2 10 log KTB 0.001 C1dBm2 10 log C1watts2 0.001 99
- 105. Digital Modulation (38) where N0 noise power density (watts per hertz) N thermal noise power (watts) B bandwidth (hertz) Stated in dBm, (39) N(dBm) 10 log B (40) Combining Equations 29 and 38 yields (41) Stated in dBm, (42) Energy per bit-to-noise power density ratio is used to compare two or more digital modulation systems that use different transmission rates (bit rates), modulation schemes (FSK, PSK, QAM), or encoding techniques (M-ary). The energy per bit-to-noise power density ratio is simply the ratio of the energy of a single bit to the noise power present in 1 Hz of bandwidth. Thus, Eb/N0 normalizes all multiphase modulation schemes to a com- mon noise bandwidth, allowing for a simpler and more accurate comparison of their error performance. Mathematically, Eb/N0 is (43) where EbN0 is the energy per bit-to-noise power density ratio. Rearranging Equation 43 yields the following expression: (44) where Eb/N0 energy per bit-to-noise power density ratio C/N carrier-to-noise power ratio B/fb noise bandwidth-to-bit rate ratio Stated in dB, (45) or 10 log Eb 10 log N0 (46) From Equation 44, it can be seen that the Eb/N0 ratio is simply the product of the carrier-to- noise power ratio and the noise bandwidth-to-bit rate ratio. Also, from Equation 44, it can be seen that when the bandwidth equals the bit rate, Eb/N0 C/N. In general, the minimum carrier-to-noise power ratio required for QAM systems is less than that required for comparable PSK systems, Also, the higher the level of encoding used (the higher the value of M), the higher the minimum carrier-to-noise power ratio. Example 15 For a QPSK system and the given parameters, determine a. Carrier power in dBm. b. Noise power in dBm. Eb N0 1dB2 10 log C N 10 log B fb Eb N0 C N B fb Eb N0 Cfb NB CB Nfb N01dBm2 10 log K 0.001 10 log T N0 KTB B KT 1WHz2 N01dBm2 10 log N 0.001 10 log B N0 N B 1WHz2 100
- 106. Digital Modulation c. Noise power density in dBm. d. Energy per bit in dBJ. e. Carrier-to-noise power ratio in dB. f. Eb/N0 ratio. Solution a. The carrier power in dBm is determined by substituting into Equation 28: b. The noise power in dBm is determined by substituting into Equation 30: c. The noise power density is determined by substituting into Equation 40: N0 109.2 dBm 10 log 120 kHz 160 dBm d. The energy per bit is determined by substituting into Equation 36: e. The carrier-to-noise power ratio is determined by substituting into Equation 34: f. The energy per bit-to-noise density ratio is determined by substituting into Equation 45: 13 ERROR PERFORMANCE 13-1 PSK Error Performance The bit error performance for the various multiphase digital modulation systems is directly re- lated to the distance between points on a signal state-space diagram. For example, on the signal state-space diagram for BPSK shown in Figure 47a, it can be seen that the two signal points (logic 1 and logic 0) have maximum separation (d) for a given power level (D). In essence, one BPSK signal state is the exact negative of the other. As the figure shows, a noise vector (VN), when combined with the signal vector (VS), effectively shifts the phase of the signaling element (VSE) alpha degrees. If the phase shift exceeds 90°, the signal element is shifted beyond the threshold points into the error region. For BPSK, it would require a noise vector of sufficient am- plitude and phase to produce more than a 90° phase shift in the signaling element to produce an error. For PSK systems, the general formula for the threshold points is (47) where M is the number of signal states. The phase relationship between signaling elements for BPSK (i.e., 180° out of phase) is the optimum signaling format, referred to as antipodal signaling, and occurs only when two binary signal levels are allowed and when one signal is the exact negative of the other. Because no other bit-by-bit signaling scheme is any better, antipodal performance is often used as a reference for comparison. The error performance of the other multiphase PSK systems can be compared with that of BPSK simply by determining the relative decrease in error distance between points TP ± π M Eb N0 19.2 10 log 120 kHz 60 kbps 22.2 dB C N 10 log 1012 1.2 1014 19.2 dB Eb 10 log 1012 60 kbps 167.8 dBJ N 10 log 1.2 1014 0.001 109.2 dBm C 10 log 1012 0.001 90 dBm C 1012 W fb 60 kbps N 1.2 1014 W B 120 kHz 101
- 107. Digital Modulation FIGURE 47 PSK error region: (a) BPSK; (b) QPSK on a signal state-space diagram. For PSK, the general formula for the maximum distance between signaling points is given by (48) where d error distance M number of phases D peak signal amplitude Rearranging Equation 48 and solving for d yields (49) Figure 47b shows the signal state-space diagram for QPSK. From Figure 47 and Equa- tion 48, it can be seen that QPSK can tolerate only a 45° phase shift. From Equation 47, d ¢2 sin 180° M ≤ D sin θ sin 360° 2M d2 D 102
- 108. Digital Modulation the maximum phase shift for 8-PSK and 16-PSK is 22.5° and 11.25°, respectively. Con- sequently, the higher levels of modulation (i.e., the greater the value of M) require a greater energy per bit-to-noise power density ratio to reduce the effect of noise interference. Hence, the higher the level of modulation, the smaller the angular separation between signal points and the smaller the error distance. The general expression for the bit error probability of an M-phase PSK system is (50) where erf error function By substituting into Equation 50, it can be shown that QPSK provides the same error performance as BPSK. This is because the 3-dB reduction in error distance for QPSK is off- set by the 3-dB decrease in its bandwidth (in addition to the error distance, the relative widths of the noise bandwidths must also be considered). Thus, both systems provide opti- mum performance. Figure 48 shows the error performance for 2-, 4-, 8-, 16-, and 32-PSK sys- tems as a function of Eb/N0. z sin1πM2 12log2M2 12EbN0 2 P1e2 1 log2M erf1z2 FIGURE 48 Error rates of PSK modulation systems 103
- 109. Digital Modulation Example 16 Determine the minimum bandwidth required to achieve a P(e) of 107 for an 8-PSK system operat- ing at 10 Mbps with a carrier-to-noise power ratio of 11.7 dB. Solution From Figure 48, the minimum Eb/N0 ratio to achieve a P(e) of 107 for an 8-PSK system is 14.7 dB. The minimum bandwidth is found by rearranging Equation 44: B 2 10 Mbps 20 MHz 13-2 QAM Error Performance For a large number of signal points (i.e., M-ary systems greater than 4), QAM outperforms PSK. This is because the distance between signaling points in a PSK system is smaller than the distance between points in a comparable QAM system. The general expression for the distance between adjacent signaling points for a QAM system with L levels on each axis is (51) where d error distance L number of levels on each axis D peak signal amplitude In comparing Equation 49 to Equation 51, it can be seen that QAM systems have an advantage over PSK systems with the same peak signal power level. The general expression for the bit error probability of an L-level QAM system is (52) where erfc(z) is the complementary error function. Figure 49 shows the error performance for 4-, 16-, 32-, and 64-QAM systems as a function of Eb/N0. Table 4 lists the minimum carrier-to-noise power ratios and energy per bit-to-noise power density ratios required for a probability of error 106 for several PSK and QAM modulation schemes. Which system requires the highest Eb/N0 ratio for a probability of error of 106 , a four-level QAM system or an 8-PSK system? Solution From Figure 49, the minimum Eb/N0 ratio required for a four-level QAM system is 10.6 dB. From Figure 48, the minimum Eb/N0 ratio required for an 8-PSK system is 14 dB. Therefore, to achieve a P(e) of 106 , a four-level QAM system would require 3.4 dB less Eb/N0 ratio. z 2log2L L 1 B Eb N0 P1e2 1 log2L ¢ L 1 L ≤ erfc1z2 d 22 L 1 D B fb antilog 3 2 14.7 dB 11.7 dB 3 dB B fb Eb N0 C N Example 17 104
- 110. Digital Modulation FIGURE 49 Error rates of QAM modulation systems Table 4 Performance Comparison of Various Digital Modulation Schemes (BER = 106 ) Modulation Technique C/N Ratio (dB) Eb/N0 Ratio (dB) BPSK 10.6 10.6 QPSK 13.6 10.6 4-QAM 13.6 10.6 8-QAM 17.6 10.6 8-PSK 18.5 14 16-PSK 24.3 18.3 16-QAM 20.5 14.5 32-QAM 24.4 17.4 64-QAM 26.6 18.8 105
- 111. Digital Modulation FIGURE 50 Error rates for FSK modulation systems 13-3 FSK Error Performance The error probability for FSK systems is evaluated in a somewhat different manner than PSK and QAM. There are essentially only two types of FSK systems: noncoherent (asyn- chronous) and coherent (synchronous). With noncoherent FSK, the transmitter and receiver are not frequency or phase synchronized. With coherent FSK, local receiver reference sig- nals are in frequency and phase lock with the transmitted signals. The probability of error for noncoherent FSK is (53) The probability of error for coherent FSK is (54) Figure 50 shows probability of error curves for both coherent and noncoherent FSK for sev- eral values of Eb/N0. From Equations 53 and 54, it can be determined that the probability of error for noncoherent FSK is greater than that of coherent FSK for equal energy per bit- to-noise power density ratios. QUESTIONS 1. Explain digital transmission and digital radio. 2. Define information capacity. 3. What are the three most predominant modulation schemes used in digital radio systems? P1e2 erfc B Eb N0 P1e2 1 2 exp ¢ Eb 2N0 ≤ 106
- 112. Digital Modulation 4. Explain the relationship between bits per second and baud for an FSK system. 5. Define the following terms for FSK modulation: frequency deviation, modulation index, and dev- iation ratio. 6. Explain the relationship between (a) the minimum bandwidth required for an FSK system and the bit rate and (b) the mark and space frequencies. 7. What is the difference between standard FSK and MSK? What is the advantage of MSK? 8. Define PSK. 9. Explain the relationship between bits per second and baud for a BPSK system. 10. What is a constellation diagram, and how is it used with PSK? 11. Explain the relationship between the minimum bandwidth required for a BPSK system and the bit rate. 12. Explain M-ary. 13. Explain the relationship between bits per second and baud for a QPSK system. 14. Explain the significance of the I and Q channels in a QPSK modulator. 15. Define dibit. 16. Explain the relationship between the minimum bandwidth required for a QPSK system and the bit rate. 17. What is a coherent demodulator? 18. What advantage does OQPSK have over conventional QPSK? What is a disadvantage of OQPSK? 19. Explain the relationship between bits per second and baud for an 8-PSK system. 20. Define tribit. 21. Explain the relationship between the minimum bandwidth required for an 8-PSK system and the bit rate. 22. Explain the relationship between bits per second and baud for a 16-PSK system. 23. Define quadbit. 24. Define QAM. 25. Explain the relationship between the minimum bandwidth required for a 16-QAM system and the bit rate. 26. What is the difference between PSK and QAM? 27. Define bandwidth efficiency. 28. Define carrier recovery. 29. Explain the differences between absolute PSK and differential PSK. 30. What is the purpose of a clock recovery circuit? When is it used? 31. What is the difference between probability of error and bit error rate? PROBLEMS 1. Determine the bandwidth and baud for an FSK signal with a mark frequency of 32 kHz, a space frequency of 24 kHz, and a bit rate of 4 kbps. 2. Determine the maximum bit rate for an FSK signal with a mark frequency of 48 kHz, a space fre- quency of 52 kHz, and an available bandwidth of 10 kHz. 3. Determine the bandwidth and baud for an FSK signal with a mark frequency of 99 kHz, a space frequency of 101 kHz, and a bit rate of 10 kbps. 4. Determine the maximum bit rate for an FSK signal with a mark frequency of 102 kHz, a space frequency of 104 kHz, and an available bandwidth of 8 kHz. 5. Determine the minimum bandwidth and baud for a BPSK modulator with a carrier frequency of 40 MHz and an input bit rate of 500 kbps. Sketch the output spectrum. 6. For the QPSK modulator shown in Figure 17, change the 90° phase-shift network to 90° and sketch the new constellation diagram. 7. For the QPSK demodulator shown in Figure 21, determine the I and Q bits for an input signal of sin ωct cos ωct. 107
- 113. 8. For an 8-PSK modulator with an input data rate (fb) equal to 20 Mbps and a carrier frequency of 100 MHz, determine the minimum double-sided Nyquist bandwidth (fN) and the baud. Sketch the output spectrum. 9. For the 8-PSK modulator shown in Figure 23, change the reference oscillator to cos ωct and sketch the new constellation diagram. 10. For a 16-QAM modulator with an input bit rate (fb) equal to 20 Mbps and a carrier frequency of 100 MHz, determine the minimum double-sided Nyquist bandwidth (fN) and the baud. Sketch the output spectrum. 11. For the 16-QAM modulator shown in Figure 33, change the reference oscillator to cos ωct and determine the output expressions for the following I, I , Q, and Q input conditions: 0000, 1111, 1010, and 0101. 12. Determine the bandwidth efficiency for the following modulators: a. QPSK, fb 10 Mbps b. 8-PSK, fb 21 Mbps c. 16-QAM, fb 20 Mbps 13. For the DBPSK modulator shown in Figure 40a, determine the output phase sequence for the fol- lowing input bit sequence: 00110011010101 (assume that the reference bit 1). 14. For a QPSK system and the given parameters, determine a. Carrier power in dBm. b. Noise power in dBm. c. Noise power density in dBm. d. Energy per bit in dBJ. e. Carrier-to-noise power ratio. f. Eb/N0 ratio. C 1013 W fb 30 kbps N 0.06 1015 W B 60 kHz 15. Determine the minimum bandwidth required to achieve a P(e) of 106 for an 8-PSK system op- erating at 20 Mbps with a carrier-to-noise power ratio of 11 dB. 16. Determine the minimum bandwidth and baud for a BPSK modulator with a carrier frequency of 80 MHz and an input bit rate fb 1 Mbps. Sketch the output spectrum. 17. For the QPSK modulator shown in Figure 17, change the reference oscillator to cos ωct and sketch the new constellation diagram. 18. For the QPSK demodulator shown in Figure 21, determine the I and Q bits for an input signal sin ωct cos ωct. 19. For an 8-PSK modulator with an input bit rate fb 10 Mbps and a carrier frequency fc 80 MHz, determine the minimum Nyquist bandwidth and the baud. Sketch the output spectrum. 20. For the 8-PSK modulator shown in Figure 23, change the 90° phase-shift network to a 90° phase shifter and sketch the new constellation diagram. 21. For a 16-QAM modulator with an input bit rate fb 10 Mbps and a carrier frequency fc 60 MHz, determine the minimum double-sided Nyquist frequency and the baud. Sketch the output spectrum. 22. For the 16-QAM modulator shown in Figure 33, change the 90° phase shift network to a 90° phase shifter and determine the output expressions for the following I, I , Q, and Q input condi- tions: 0000, 1111, 1010, and 0101. 23. Determine the bandwidth efficiency for the following modulators: a. QPSK, fb 20 Mbps b. 8-PSK, fb 28 Mbps c. 16-PSK, fb 40 Mbps 24. For the DBPSK modulator shown in Figure 40a, determine the output phase sequence for the fol- lowing input bit sequence: 11001100101010 (assume that the reference bit is a logic 1). Digital Modulation 108
- 114. ANSWERS TO SELECTED PROBLEMS 1. 16 kHz, 4000 baud 3. 22 kHz, 10 kbaud 5. 5 MHz, 5 Mbaud 7. I 1, Q 0 9. 11. 13. Input 00110011010101 XNOR 101110111001100 15. 40 MHz 17. 19. 3.33 MHz, 3.33 Mbaud 21. 2.5 MHz, 2.5 Mbaud 23. a. 2 bps/Hz b. 3 bps/Hz c. 4 bps/Hz Digital Modulation Q I Phase 0 0 135° 0 1 45° 1 0 135° 1 1 45° Q Q’ I I’ Phase 0 0 0 0 45° 1 1 1 1 135° 1 0 1 0 135° 0 1 0 1 45° Q I C Phase 0 0 0 22.5° 0 0 1 67.5° 0 1 0 22.5° 0 1 1 67.5° 1 0 0 157.5° 1 0 1 112.5° 1 1 0 157.5° 1 1 1 112.5° 109
- 115. 110
- 116. Introduction to Data Communications and Networking CHAPTER OUTLINE 1 Introduction 6 Open Systems Interconnection 2 History of Data Communications 7 Data Communications Circuits 3 Data Communications Network Architecture, 8 Serial and Parallel Data Transmission Protocols, and Standards 9 Data Communications Circuit Arrangements 4 Standards Organizations for Data 10 Data Communications Networks Communications 11 Alternate Protocol Suites 5 Layered Network Architecture OBJECTIVES ■ Define the following terms: data, data communications, data communications circuit, and data communications network ■ Give a brief description of the evolution of data communications ■ Define data communications network architecture ■ Describe data communications protocols ■ Describe the basic concepts of connection-oriented and connectionless protocols ■ Describe syntax and semantics and how they relate to data communications ■ Define data communications standards and explain why they are necessary ■ Describe the following standards organizations: ISO, ITU-T, IEEE, ANSI, EIA, TIA, IAB, ETF, and IRTF ■ Define open systems interconnection ■ Name and explain the functions of each of the layers of the seven-layer OSI model ■ Define station and node ■ Describe the fundamental block diagram of a two-station data communications circuit and explain how the fol- lowing terms relate to it: source, transmitter, transmission medium, receiver, and destination ■ Describe serial and parallel data transmission and explain the advantages and disadvantages of both types of trans- missions From Chapter 3 of Advanced Electronic Communications Systems, Sixth Edition. Wayne Tomasi. Copyright © 2004 by Pearson Education, Inc. Published by Pearson Prentice Hall. All rights reserved. 111
- 117. ■ Define data communications circuit arrangements ■ Describe the following transmission modes: simplex, half duplex, full duplex, and full/full duplex ■ Define data communications network ■ Describe the following network components, functions, and features: servers, clients, transmission media, shared data, shared printers, and network interface card ■ Define local operating system ■ Define network operating system ■ Describe peer-to-peer client/server and dedicated client/server networks ■ Define network topology and describe the following: star, bus, ring, mesh, and hybrid ■ Describe the following classifications of networks: LAN, MAN, WAN, GAN, building backbone, campus back- bone, and enterprise network ■ Briefly describe the TCP/IP hierarchical model ■ Briefly describe the Cisco three-layer hierarchical model 1 INTRODUCTION Since the early 1970s, technological advances around the world have occurred at a phe- nomenal rate, transforming the telecommunications industry into a highly sophisticated and extremely dynamic field. Where previously telecommunications systems had only voice to accommodate, the advent of very large-scale integration chips and the accom- panying low-cost microprocessors, computers, and peripheral equipment has dramati- cally increased the need for the exchange of digital information. This, of course, neces- sitated the development and implementation of higher-capacity and much faster means of communicating. In the data communications world, data generally are defined as information that is stored in digital form. The word data is plural; a single unit of data is a datum. Data com- munications is the process of transferring digital information (usually in binary form) be- tween two or more points. Information is defined as knowledge or intelligence. Information that has been processed, organized, and stored is called data. The fundamental purpose of a data communications circuit is to transfer digital in- formation from one place to another. Thus, data communications can be summarized as the transmission, reception, and processing of digital information. The original source infor- mation can be in analog form, such as the human voice or music, or in digital form, such as binary-coded numbers or alphanumeric codes. If the source information is in analog form, it must be converted to digital form at the source and then converted back to analog form at the destination. A network is a set of devices (sometimes called nodes or stations) interconnected by media links. Data communications networks are systems of interrelated computers and computer equipment and can be as simple as a personal computer connected to a printer or two personal computers connected together through the public telephone network. On the other hand, a data communications network can be a complex com- munications system comprised of one or more mainframe computers and hundreds, thousands, or even millions of remote terminals, personal computers, and worksta- tions. In essence, there is virtually no limit to the capacity or size of a data communi- cations network. Years ago, a single computer serviced virtually every computing need. Today, the single-computer concept has been replaced by the networking concept, where a large num- ber of separate but interconnected computers share their resources. Data communications networks and systems of networks are used to interconnect virtually all kinds of digital computing equipment, from automatic teller machines (ATMs) to bank computers; per- sonal computers to information highways, such as the Internet; and workstations to main- Introduction to Data Communications and Networking 112
- 118. frame computers. Data communications networks can also be used for airline and hotel reservation systems, mass media and news networks, and electronic mail delivery systems. The list of applications for data communications networks is virtually endless. 2 HISTORY OF DATA COMMUNICATIONS It is highly likely that data communications began long before recorded time in the form of smoke signals or tom-tom drums, although they surely did not involve electricity or an elec- tronic apparatus, and it is highly unlikely that they were binary coded. One of the earliest means of communicating electrically coded information occurred in 1753, when a proposal submitted to a Scottish magazine suggested running a communications line between vil- lages comprised of 26 parallel wires, each wire for one letter of the alphabet. A Swiss in- ventor constructed a prototype of the 26-wire system, but current wire-making technology proved the idea impractical. In 1833, Carl Friedrich Gauss developed an unusual system based on a five-by-five matrix representing 25 letters (I and J were combined). The idea was to send messages over a single wire by deflecting a needle to the right or left between one and five times. The ini- tial set of deflections indicated a row, and the second set indicated a column. Consequently, it could take as many as 10 deflections to convey a single character through the system. If we limit the scope of data communications to methods that use binary-coded electri- cal signals to transmit information, then the first successful (and practical) data communica- tions system was invented by Samuel F. B. Morse in 1832 and called the telegraph. Morse also developed the first practical data communications code, which he called the Morse code. With telegraph, dots and dashes (analogous to logic 1s and 0s) are transmitted across a wire using electromechanical induction.Various combinations of dots, dashes, and pauses represented bi- nary codes for letters, numbers, and punctuation marks. Because all codes did not contain the same number of dots and dashes, Morse’s system combined human intelligence with electron- ics, as decoding was dependent on the hearing and reasoning ability of the person receiving the message. (Sir CharlesWheatstone and SirWilliam Cooke allegedly invented the first telegraph in England, but their contraption required six different wires for a single telegraph line.) In 1840, Morse secured anAmerican patent for the telegraph, and in 1844 the first tele- graph line was established between Baltimore and Washington, D.C., with the first message conveyed over this system being “What hath God wrought!” In 1849, the first slow-speed telegraph printer was invented, but it was not until 1860 that high-speed (15-bps) printers were available. In 1850, Western Union Telegraph Company was formed in Rochester, New York, for the purpose of carrying coded messages from one person to another. In 1874, Emile Baudot invented a telegraph multiplexer, which allowed signals from up to six different telegraph machines to be transmitted simultaneously over a single wire. The telephone was invented in 1875 by Alexander Graham Bell and, unfortunately, very lit- tle new evolved in telegraph until 1899, when Guglielmo Marconi succeeded in sending ra- dio (wireless) telegraph messages. Telegraph was the only means of sending information across large spans of water until 1920, when the first commercial radio stations carrying voice information were installed. It is unclear exactly when the first electrical computer was developed. Konrad Zuis, a German engineer, demonstrated a computing machine sometime in the late 1930s; how- ever, at the time, Hitler was preoccupied trying to conquer the rest of the world, so the proj- ect fizzled out. Bell Telephone Laboratories is given credit for developing the first special- purpose computer in 1940 using electromechanical relays for performing logical operations. However, J. Presper Eckert and John Mauchley at the University of Pennsylva- nia are given credit by some for beginning modern-day computing when they developed the ENIAC computer on February 14, 1946. Introduction to Data Communications and Networking 113
- 119. In 1949, the U.S. National Bureau of Standards developed the first all-electronic diode-based computer capable of executing stored programs. The U.S. Census Bureau in- stalled the machine, which is considered the first commercially produced American com- puter. In the 1950s, computers used punch cards for inputting information, printers for outputting information, and magnetic tape reels for permanently storing information. These early computers could process only one job at a time using a technique called batch processing. The first general-purpose computer was an automatic sequence-controlled calculator developed jointly by Harvard University and International Business Machines (IBM) Cor- poration. The UNIVAC computer, built in 1951 by Remington Rand Corporation, was the first mass-produced electronic computer. In the 1960s, batch-processing systems were replaced by on-line processing systems with terminals connected directly to the computer through serial or parallel communica- tions lines. The 1970s introduced microprocessor-controlled microcomputers, and by the 1980s personal computers became an essential item in the home and workplace. Since then, the number of mainframe computers, small business computers, personal computers, and computer terminals has increased exponentially, creating a situation where more and more people have the need (or at least think they have the need) to exchange digital information with each other. Consequently, the need for data communications circuits, networks, and systems has also increased exponentially. Soon after the invention of the telephone, the American Telephone and Telegraph Company (ATT) emerged, providing both long-distance and local telephone service and data communications service throughout the United States. The vast ATT system was referred to by some as the “Bell System” and by others as “Ma Bell.” During this time, Western Union Corporation provided telegraph service. Until 1968, the ATT op- erating tariff allowed only equipment furnished by ATT to be connected to ATT lines. In 1968, a landmark Supreme Court decision, the Carterfone decision, allowed non-Bell companies to interconnect to the vast ATT communications network. This decision started the interconnect industry, which has led to competitive data communi- cations offerings by a large number of independent companies. In 1983, as a direct re- sult of an antitrust suit filed by the federal government, ATT agreed in a court settle- ment to divest itself of operating companies that provide basic local telephone service to the various geographic regions of the United States. Since the divestiture, the com- plexity of the public telephone system in the United States has grown even more in- volved and complicated. Recent developments in data communications networking, such as the Internet, in- tranets, and the World Wide Web (WWW), have created a virtual explosion in the data com- munications industry. A seemingly infinite number of people, from homemaker to chief ex- ecutive officer, now feel a need to communicate over a finite number of facilities. Thus, the demand for higher-capacity and higher-speed data communications systems is increasing daily with no end in sight. The Internet is a public data communications network used by millions of people all over the world to exchange business and personal information. The Internet began to evolve in 1969 at the Advanced Research Projects Agency (ARPA). ARPANET was formed in the late 1970s to connect sites around the United States. From the mid-1980s to April 30, 1995, the National Science Foundation (NSF) funded a high-speed backbone called NSFNET. Intranets are private data communications networks used by many companies to ex- change information among employees and resources. Intranets normally are used for secu- rity reasons or to satisfy specific connectivity requirements. Company intranets are gener- ally connected to the public Internet through a firewall, which converts the intranet addressing system to the public Internet addressing system and provides security function- ality by filtering incoming and outgoing traffic based on addressing and protocols. Introduction to Data Communications and Networking 114
- 120. The World Wide Web (WWW) is a server-based application that allows subscribers to access the services offered by the Web. Browsers, such as Netscape Communicator and Mi- crosoft Internet Explorer, are commonly used for accessing data over the WWW. 3 DATA COMMUNICATIONS NETWORK ARCHITECTURE, PROTOCOLS, AND STANDARDS 3-1 Data Communications Network Architecture A data communications network is any system of computers, computer terminals, or com- puter peripheral equipment used to transmit and/or receive information between two or more locations. Network architectures outline the products and services necessary for the individual components within a data communications network to operate together. In essence, network architecture is a set of equipment, transmission media, and proce- dures that ensures that a specific sequence of events occurs in a network in the proper order to produce the intended results. Network architecture must include sufficient information to allow a program or a piece of hardware to perform its intended function. The primary goal of network architecture is to give the users of the network the tools necessary for setting up the network and performing data flow control.A network architecture outlines the way in which a data communications network is arranged or structured and generally includes the concept of levels or layers of functional responsibility within the architecture. The functional respon- sibilities include electrical specifications, hardware arrangements, and software procedures. Networks and network protocols fall into three general classifications: current, legacy, and legendary. Current networks include the most modern and sophisticated net- works and protocols available. If a network or protocol becomes a legacy, no one really wants to use it, but for some reason it just will not go away. When an antiquated network or protocol finally disappears, it becomes legendary. In general terms, computer networks can be classified in two different ways: broadcast and point to point. With broadcast networks, all stations and devices on the network share a single communications channel. Data are propagated through the network in relatively short messages sometimes called frames, blocks, or packets. Many or all subscribers of the net- work receive transmitted messages, and each message contains an address that identifies specifically which subscriber (or subscribers) is intended to receive the message.When mes- sages are intended for all subscribers on the network, it is called broadcasting, and when messages are intended for a specific group of subscribers, it is called multicasting. Point-to-point networks have only two stations. Therefore, no addresses are needed. All transmissions from one station are intended for and received by the other station. With point-to-point networks, data are often transmitted in long, continuous messages, some- times requiring several hours to send. In more specific terms, point-to-point and broadcast networks can be subdivided into many categories in which one type of network is often included as a subnetwork of another. 3-2 Data Communications Protocols Computer networks communicate using protocols, which define the procedures that the sys- tems involved in the communications process will use. Numerous protocols are used today to provide networking capabilities, such as how much data can be sent, how it will be sent, how it willbeaddressed,andwhatprocedurewillbeusedtoensurethattherearenoundetectederrors. Protocols are arrangements between people or processes. In essence, a protocol is a set of customs, rules, or regulations dealing with formality or precedence, such as diplomatic or military protocol. Each functional layer of a network is responsible for providing a spe- cific service to the data being transported through the network by providing a set of rules, called protocols, that perform a specific function (or functions) within the network. Data communications protocols are sets of rules governing the orderly exchange of data within Introduction to Data Communications and Networking 115
- 121. the network or a portion of the network, whereas network architecture is a set of layers and protocols that govern the operation of the network. The list of protocols used by a system is called a protocol stack, which generally includes only one protocol per layer. Layered net- work architectures consist of two or more independent levels. Each level has a specific set of responsibilities and functions, including data transfer, flow control, data segmentation and reassembly, sequence control, error detection and correction, and notification. 3-2-1 Connection-oriented and connectionless protocols. Protocols can be gen- erally classified as either connection oriented or connectionless. With a connection-ori- ented protocol, a logical connection is established between the endpoints (e.g., a virtual cir- cuit) prior to the transmission of data. Connection-oriented protocols operate in a manner similar to making a standard telephone call where there is a sequence of actions and ac- knowledgments, such as setting up the call, establishing the connection, and then discon- necting. The actions and acknowledgments include dial tone, Touch-Tone signaling, ring- ing and ring-back signals, and busy signals. Connection-oriented protocols are designed to provide a high degree of reliability for data moving through the network. This is accomplished by using a rigid set of procedures for establishing the connection, transferring the data, acknowledging the data, and then clearing the connection. In a connection-oriented system, each packet of data is assigned a unique sequence number and an associated acknowledgement number to track the data as they travel through a network. If data are lost or damaged, the destination station requests that they be re-sent. A connection-oriented protocol is depicted in Figure 1a. Characteris- tics of connection-oriented protocols include the following: 1. A connection process called a handshake occurs between two stations before any data are actually transmitted. Connections are sometimes referred to as sessions, virtual circuits, or logical connections. 2. Most connection-oriented protocols require some means of acknowledging the data as they are being transmitted. Protocols that use acknowledgment procedures provide a high level of network reliability. 3. Connection-oriented protocols often provide some means of error control (i.e., er- ror detection and error correction). Whenever data are found to be in error, the re- ceiving station requests a retransmission. 4. When a connection is no longer needed, a specific handshake drops the connection. Connectionless protocols are protocols where data are exchanged in an unplanned fashion without prior coordination between endpoints (e.g., a datagram). Connectionless protocols do not provide the same high degree of reliability as connection-oriented proto- cols; however, connectionless protocols offer a significant advantage in transmission speed. Connectionless protocols operate in a manner similar to the U.S. Postal Service, where in- formation is formatted, placed in an envelope with source and destination addresses, and then mailed. You can only hope the letter arrives at its destination. A connectionless proto- col is depicted in Figure 1b. Characteristics of connectionless protocols are as follow: 1. Connectionless protocols send data with a source and destination address without a handshake to ensure that the destination is ready to receive the data. 2. Connectionless protocols usually do not support error control or acknowledgment procedures, making them a relatively unreliable method of data transmission. 3. Connectionless protocols are used because they are often more efficient, as the data being transmitted usually do not justify the extra overhead required by connection-oriented protocols. Introduction to Data Communications and Networking 116
- 122. Data acknowledgment Data transmitted Setup response Setup request Connection clear request NETWORK Station 1 Station 2 Clear response (a) Data Data Data Data Data (b) NETWORK Station 1 Station 2 FIGURE 1 Network protocols: (a) connection-oriented; (b) connectionless 3-2-2 Syntax and semantics. Protocols include the concepts of syntax and se- mantics. Syntax refers to the structure or format of the data within the message, which in- cludes the sequence in which the data are sent. For example, the first byte of a message might be the address of the source and the second byte the address of the destination. Se- mantics refers to the meaning of each section of data. For example, does a destination ad- dress identify only the location of the final destination, or does it also identify the route the data takes between the sending and receiving locations? 3-3 Data Communications Standards During the past several decades, the data communications industry has grown at an astro- nomical rate. Consequently, the need to provide communications between dissimilar com- puter equipment and systems has also increased. A major issue facing the data communi- cations industry today is worldwide compatibility. Major areas of interest are software and programming language, electrical and cable interface, transmission media, communica- tions signal, and format compatibility. Thus, to ensure an orderly transfer of information, it has been necessary to establish standard means of governing the physical, electrical, and procedural arrangements of a data communications system. A standard is an object or procedure considered by an authority or by general consent as a basis of comparison. Standards are authoritative principles or rules that imply a model or pattern for guidance by comparison. Data communications standards are guidelines that Introduction to Data Communications and Networking 117
- 123. have been generally accepted by the data communications industry. The guidelines outline procedures and equipment configurations that help ensure an orderly transfer of informa- tion between two or more pieces of data communications equipment or two or more data communications networks. Data communications standards are not laws, however—they are simply suggested ways of implementing procedures and accomplishing results. If everyone complies with the standards, everyone’s equipment, procedures, and processes will be compatible with everyone else’s, and there will be little difficulty communicating information through the system. Today, most companies make their products to comply with standards. There are two basic types of standards: proprietary (closed) system and open sys- tem. Proprietary standards are generally manufactured and controlled by one company. Other companies are not allowed to manufacture equipment or write software using this standard. An example of a proprietary standard is Apple Macintosh computers. Advan- tages of proprietary standards are tighter control, easier consensus, and a monopoly. Disadvantages include lack of choice for the customers, higher financial investment, overpricing, and reduced customer protection against the manufacturer going out of business. With open system standards, any company can produce compatible equipment or software; however, often a royalty must be paid to the original company. An example of an open system standard is IBM’s personal computer. Advantages of open system standards are customer choice, compatibility between venders, and competition by smaller compa- nies. Disadvantages include less product control and increased difficulty acquiring agree- ment between vendors for changes or updates. In addition, standard items are not always as compatible as we would like them to be. 4 STANDARDS ORGANIZATIONS FOR DATA COMMUNICATIONS Aconsortium of organizations, governments, manufacturers, and users meet on a regular ba- sis to ensure an orderly flow of information within data communications networks and sys- tems by establishing guidelines and standards. The intent is that all data communications equipment manufacturers and users comply with these standards. Standards organizations generate, control, and administer standards. Often, competing companies will form a joint committee to create a compromised standard that is acceptable to everyone.The most promi- nent organizations relied on in North America to publish standards and make recommenda- tions for the data, telecommunications, and networking industries are shown in Figure 2. 4-1 International Standards Organization (ISO) Created in 1946, the International Standards Organization (ISO) is the international or- ganization for standardization on a wide range of subjects. The ISO is a voluntary, nontreaty organization whose membership is comprised mainly of members from the standards com- mittees of various governments throughout the world. The ISO creates the sets of rules and standards for graphics and document exchange and provides models for equipment and sys- tem compatibility, quality enhancement, improved productivity, and reduced costs. The ISO is responsible for endorsing and coordinating the work of the other standards organi- zations. The member body of the ISO from the United States is theAmerican National Stan- dards Institute (ANSI). 4-2 International Telecommunications Union— Telecommunications Sector TheInternationalTelecommunicationsUnion—TelecommunicationsSector(ITU-T),formerly theComitéConsultatifInternationaledeTélégraphieetTéléphonie(CCITT),isoneoffourper- manent parts of the International Telecommunications Union based in Geneva, Switzerland. Introduction to Data Communications and Networking 118
- 124. ITU-T IEEE TIA EIA ISO ANSI IETF IAB IRTF FIGURE 2 Standards organizations for data and network communications Membership in the ITU-T consists of government authorities and representatives from many countries. The ITU-T is now the standards organization for the United Nations and develops the recommended sets of rules and standards for telephone and data communications.The ITU-T has developed three sets of specifications: the V series for modem interfacing and data trans- mission over telephone lines; the X series for data transmission over public digital networks, e-mail, and directory services; and the I and Q series for Integrated Services Digital Network (ISDN)anditsextensionBroadbandISDN(sometimescalledtheInformationSuperhighway). The ITU-T is separated into 14 study groups that prepare recommendations on the following topics: Network and service operation Tariff and accounting principles Telecommunications management network and network maintenance Protection against electromagnetic environment effects Outside plant Data networks and open system communications Characteristics of telematic systems Television and sound transmission Language and general software aspects for telecommunications systems Signaling requirements and protocols End-to-end transmission performance of networks and terminals General network aspects Transport networks, systems, and equipment Multimedia services and systems 4-3 Institute of Electrical and Electronics Engineers The Institute of Electrical and Electronics Engineers (IEEE) is an international professional organization founded in the United States and is comprised of electronics, computer, and communications engineers. The IEEE is currently the world’s largest professional society Introduction to Data Communications and Networking 119
- 125. with over 200,000 members. The IEEE works closely with ANSI to develop communica- tions and information processing standards with the underlying goal of advancing theory, creativity, and product quality in any field associated with electrical engineering. 4-4 American National Standards Institute The American National Standards Institute (ANSI) is the official standards agency for the UnitedStatesandistheU.S.votingrepresentativefortheISO.However,ANSIisacompletely private, nonprofit organization comprised of equipment manufacturers and users of data pro- cessing equipment and services. Although ANSI has no affiliations with the federal govern- ment of the United States, it serves as the national coordinating institution for voluntary stan- dardization in the United States.ANSI membership is comprised of people from professional societies, industry associations, governmental and regulatory bodies, and consumer groups. 4-5 Electronics Industry Association The Electronics Industries Associations (EIA) is a nonprofit U.S. trade association that es- tablishes and recommends industrial standards. EIA activities include standards develop- ment, increasing public awareness, and lobbying. The EIA is responsible for developing the RS (recommended standard) series of standards for data and telecommunications. 4-6 Telecommunications Industry Association The Telecommunications Industry Association (TIA) is the leading trade association in the communications and information technology industry. The TIA facilitates business devel- opment opportunities and a competitive marketplace through market development, trade promotion, trade shows, domestic and international advocacy, and standards development. The TIA represents manufacturers of communications and information technology prod- ucts and services providers for the global marketplace through its core competencies. The TIA also facilitates the convergence of new communications networks while working for a competitive and innovative market environment. 4-7 Internet Architecture Board In 1957, theAdvanced Research ProjectsAgency (ARPA), the research arm of the Department of Defense, was created in response to the Soviet Union’s launching of Sputnik. The original purposeofARPAwastoacceleratetheadvancementoftechnologiesthatcouldpossiblybeuse- ful to the U.S. military.WhenARPANET was initiated in the late 1970s,ARPA formed a com- mittee to oversee it. In 1983, the name of the committee was changed to the Internet Activities Board(IAB).ThemeaningoftheacronymwaslaterchangedtotheInternetArchitectureBoard. Today the IAB is a technical advisory group of the Internet Society with the follow- ing responsibilities: 1. Oversees the architecture protocols and procedures used by the Internet 2. Manages the processes used to create Internet standards and serves as an appeal board for complaints of improper execution of the standardization processes 3. Is responsible for the administration of the various Internet assigned numbers 4. Acts as representative for Internet Society interests in liaison relationships with other organizations concerned with standards and other technical and organiza- tional issues relevant to the worldwide Internet 5. Acts as a source of advice and guidance to the board of trustees and officers of the Internet Society concerning technical, architectural, procedural, and policy mat- ters pertaining to the Internet and its enabling technologies 4-8 Internet Engineering Task Force The Internet Engineering Task Force (IETF) is a large international community of network designers, operators, venders, and researchers concerned with the evolution of the Internet architecture and the smooth operation of the Internet. Introduction to Data Communications and Networking 120
- 126. Source Destination Layer N + 1 Layer N + 1 Network Peer-to-peer communications Layer N + 1 to Layer N + 1 Layer N Layer N Layer N to Layer N Layer N – 1 Layer N – 1 Layer N – 1 to Layer N – 1 FIGURE 3 Peer-to-peer data communications 4-9 Internet Research Task Force The Internet Research Task Force (IRTF) promotes research of importance to the evolution of the future Internet by creating focused, long-term and small research groups working on topics related to Internet protocols, applications, architecture, and technology. 5 LAYERED NETWORK ARCHITECTURE The basic concept of layering network responsibilities is that each layer adds value to ser- vices provided by sets of lower layers. In this way, the highest level is offered the full set of services needed to run a distributed data application. There are several advantages to using a layered architecture. A layered architecture facilitates peer-to-peer communications pro- tocols where a given layer in one system can logically communicate with its corresponding layer in another system. This allows different computers to communicate at different lev- els. Figure 3 shows a layered architecture where layer N at the source logically (but not nec- essarily physically) communicates with layer N at the destination and layer N of any inter- mediate nodes. 5-1 Protocol Data Unit When technological advances occur in a layered architecture, it is easier to modify one layer’s protocol without having to modify all the other layers. Each layer is essentially in- dependent of every other layer. Therefore, many of the functions found in lower layers have been removed entirely from software tasks and replaced with hardware. The primary dis- advantage of layered architectures is the tremendous amount of overhead required. With layered architectures, communications between two corresponding layers requires a unit of data called a protocol data unit (PDU). As shown in Figure 4, a PDU can be a header added at the beginning of a message or a trailer appended to the end of a message. In a layered ar- chitecture, communications occurs between similar layers; however, data must flow through the other layers. Data flows downward through the layers in the source system and upward through the layers in the destination system. In intermediate systems, data flows up- ward first and then downward.As data passes from one layer into another, headers and trail- ers are added and removed from the PDU. The process of adding or removing PDU infor- mation is called encapsulation/decapsulation because it appears as though the PDU from the upper layer is encapsulated in the PDU from the lower layer during the downward Introduction to Data Communications and Networking 121
- 127. Layer N PDU (a) Header User information Overhead Layer N PDU (b) Trailer User information Overhead FIGURE 4 Protocol data unit: (a) header; (b) trailer System A (PDU – data) System B (PDU – data) Network Encapsulation Layer N PDU N Header Layer N + 1 PDU N + 1 Header Layer N – 1 PDU N – 1 Header Encapsulation Encapsulation Decapsulation Layer N PDU N Header Layer N + 1 PDU N + 1 Header Layer N – 1 PDU N – 1 Header Decapsulation Decapsulation FIGURE 5 Encapsulation and decapsulation movement and decapsulated during the upward movement. Encapsulate means to place in a capsule or other protected environment, and decapsulate means to remove from a capsule or other protected environment. Figure 5 illustrates the concepts of encapsulation and de- capsulation. In a layered protocol such as the one shown in Figure 3, layer N receive services from the layer immediately below it (N 1) and provides services to the layer directly above it (N 1). Layer N can provide service to more than one entity in layer N 1 by using a service access point (SAP) address to define which entity the service is intended. Introduction to Data Communications and Networking 122
- 128. Information and network information passes from one layer of a multilayered archi- tecture to another layer through a layer-to-layer interface. A layer-to-layer interface defines what information and services the lower layer must provide to the upper layer. A well-de- fined layer and layer-to-layer interface provide modularity to a network. 6 OPEN SYSTEMS INTERCONNECTION Open systems interconnection (OSI) is the name for a set of standards for communicating among computers. The primary purpose of OSI standards is to serve as a structural guide- line for exchanging information between computers, workstations, and networks. The OSI is endorsed by both the ISO and ITU-T, which have worked together to establish a set of ISO standards and ITU-T recommendations that are essentially identical. In 1983, the ISO and ITU-T (CCITT) adopted a seven-layer communications architecture reference model. Each layer consists of specific protocols for communicating. The ISO seven-layer open systems interconnection model is shown in Figure 6. This hierarchy was developed to facilitate the intercommunications of data processing equip- ment by separating network responsibilities into seven distinct layers. As with any layered architecture, overhead information is added to a PDU in the form of headers and trailers. In fact, if all seven levels of the OSI model are addressed, as little as 15% of the transmitted message is actually source information, and the rest is overhead. The result of adding head- ers to each layer is illustrated in Figure 7. Introduction to Data Communications and Networking Layer 1 Physical Transmission method used to propagate bits through a network Layer 2 Data link Frame formatting for transmitting data across a physical communications link Layer 3 Network Network addressing and packet transmission on the network Layer 4 Transport Data tracking as it moves through a network Layer 5 Session Job management tracking Layer 6 Presentation Encoding language used in transmission Layer 7 Application User networking applications and interfacing to the network ISO Layer and name Function FIGURE 6 OSI seven-layer protocol hierarchy 123
- 129. Layer 1 Physical Layer 2 Data link Layer 3 Network Layer 4 Transport Layer 5 Session Layer 6 Presentation Layer 7 Applications Host A Applications data exchange Application A Layer 1 Physical Layer 2 Data link Layer 3 Network Layer 4 Transport Layer 5 Session Layer 6 Presentation Layer 7 Applications Application A H1 H2 H3 H4 H5 H6 H7 Data H2 H3 H4 H5 H6 H7 Data H3 H4 H5 H6 H7 Data H4 H5 H6 H7 Data H5 H6 H7 Data H6 H7 Data H7 Data Data Host B System A Protocol headers (overhead) System B FIGURE 7 OSI seven-layer international protocol hierarchy. H7—applications header, H6— presentation header, H5—session header, H4—transport header, H3—network header, H2— data-link header, H1—physical header In recent years, the OSI seven-layer model has become more academic than standard, as the hierarchy does not coincide with the Internet’s four-layer protocol model. However, the basic functions of the layers are still performed, so the seven-layer model continues to serve as a reference model when describing network functions. Levels 4 to 7 address the applications aspects of the network that allow for two host computers to communicate directly. The three bottom layers are concerned with the actual mechanics of moving data (at the bit level) from one machine to another. A brief summary of the services provided by each layer is given here. 1. Physical layer. The physical layer is the lowest level of the OSI hierarchy and is responsible for the actual propagation of unstructured data bits (1s and 0s) through a transmis- Introduction to Data Communications and Networking 124
- 130. User computer Wall jack Patch panel Hub (a) User computers A B C Hub Optical fiber cable Twisted-pair cable, coax, or optical fiber User computers D E F Hub (b) FIGURE 8 OSI layer 1—physical: (a) computer-to-hub; (b) connectivity devices sion medium, which includes how bits are represented, the bit rate, and how bit synchroniza- tion is achieved. The physical layer specifies the type of transmission medium and the trans- mission mode (simplex, half duplex, or full duplex) and the physical, electrical, functional, and procedural standards for accessing data communications networks. Definitions such as con- nections, pin assignments, interface parameters, timing, maximum and minimum voltage lev- els, and circuit impedances are made at the physical level. Transmission media defined by the physical layer include metallic cable, optical fiber cable, or wireless radio-wave propagation. The physical layer for a cable connection is depicted in Figure 8a. Connectivity devices connect devices on cabled networks. An example of a connec- tivity device is a hub. A hub is a transparent device that samples the incoming bit stream and simply repeats it to the other devices connected to the hub. The hub does not examine the data to determine what the destination is; therefore, it is classified as a layer 1 compo- nent. Physical layer connectivity for a cabled network is shown in Figure 8b. The physical layer also includes the carrier system used to propagate the data signals between points in the network. Carrier systems are simply communications systems that carry data through a system using either metallic or optical fiber cables or wireless arrange- ments, such as microwave, satellites, and cellular radio systems. The carrier can use analog or digital signals that are somehow converted to a different form (encoded or modulated) by the data and then propagated through the system. 2. Data-link layer. The data-link layer is responsible for providing error-free com- munications across the physical link connecting primary and secondary stations (nodes) within a network (sometimes referred to as hop-to-hop delivery). The data-link layer pack- ages data from the physical layer into groups called blocks, frames, or packets and provides a means to activate, maintain, and deactivate the data communications link between nodes. The data-link layer provides the final framing of the information signal, provides synchro- nization, facilitates the orderly flow of data between nodes, outlines procedures for error detection and correction, and provides the physical addressing information. A block dia- gram of a network showing data transferred between two computers (A and E) at the data- link level is illustrated in Figure 9. Note that the hubs are transparent but that the switch passes the transmission on to only the hub serving the intended destination. 3. Network layer. The network layer provides details that enable data to be routed be- tween devices in an environment using multiple networks, subnetworks, or both. Network- ing components that operate at the network layer include routers and their software. The Introduction to Data Communications and Networking 125
- 131. C D Hub Switch A B Hub E I F Hub Hub G H Hub FIGURE 9 OSI layer 2—data link Hub Router Router Router Router Hub Subnet Subnets Subnet Subnet Subnet Subnet Subnet FIGURE 10 OSI layer 3—network network layer determines which network configuration is most appropriate for the function provided by the network and addresses and routes data within networks by establishing, maintaining, and terminating connections between them. The network layer provides the upper layers of the hierarchy independence from the data transmission and switching tech- nologies used to interconnect systems. It accomplishes this by defining the mechanism in which messages are broken into smaller data packets and routed from a sending node to a receiving node within a data communications network. The network layer also typically provides the source and destination network addresses (logical addresses), subnet informa- tion, and source and destination node addresses. Figure 10 illustrates the network layer of the OSI protocol hierarchy. Note that the network is subdivided into subnetworks that are separated by routers. 4. Transport layer. The transport layer controls and ensures the end-to-end integrity of the data message propagated through the network between two devices, which provides Introduction to Data Communications and Networking 126
- 132. Data Data Computer A Acknowledgment Acknowledgment Computer B Network FIGURE 11 OSI layer 4—transport Hub Server Service response Service request Client Client Client FIGURE 12 OSI layer 5—session for the reliable, transparent transfer of data between two endpoints. Transport layer re- sponsibilities includes message routing, segmenting, error recovery, and two types of basic services to an upper-layer protocol: connectionless oriented and connectionless. The trans- port layer is the highest layer in the OSI hierarchy in terms of communications and may provide data tracking, connection flow control, sequencing of data, error checking, and ap- plication addressing and identification. Figure 11 depicts data transmission at the transport layer. 5. Sessionlayer.Thesessionlayerisresponsiblefornetworkavailability(i.e.,datastor- age and processor capacity). Session layer protocols provide the logical connection entities at the application layer. These applications include file transfer protocols and sending e-mail. Session responsibilities include network log-on and log-off procedures and user authentica- tion.A session is a temporary condition that exists when data are actually in the process of be- ing transferred and does not include procedures such as call establishment, setup, or discon- nect. The session layer determines the type of dialogue available (i.e., simplex, half duplex, or full duplex). Session layer characteristics include virtual connections between applications entities,synchronizationofdataflowforrecoverypurposes,creationofdialogueunitsandac- tivity units, connection parameter negotiation, and partitioning services into functional groups. Figure 12 illustrates the establishment of a session on a data network. 6. Presentation layer. The presentation layer provides independence to the applica- tion processes by addressing any code or syntax conversion necessary to present the data to the network in a common communications format. The presentation layer specifies how end-user applications should format the data. This layer provides for translation between local representations of data and the representation of data that will be used for transfer be- tween end users. The results of encryption, data compression, and virtual terminals are ex- amples of the translation service. Introduction to Data Communications and Networking 127
- 133. Computer B Computer A Network Type Options Images JPEG, PICT, GIF Video MPEG, MIDI Data ASCII, EBCDIC FIGURE 13 OSI layer 6— presentation Networking Applications File transfer Email Printing PC Applications Database Word processing Spreadsheets FIGURE 14 OSI layer 7— applications The presentation layer translates between different data formats and protocols. Presentation functions include data file formatting, encoding, encryption and decryption of data messages, dialogue procedures, data compression algorithms, synchronization, interruption, and termination. The presentation layer performs code and character set translation (including ASCII and EBCDIC) and formatting information and determines the display mechanism for messages. Figure 13 shows an illustration of the presentation layer. 7. Application layer. The application layer is the highest layer in the hierarchy and is analogous to the general manager of the network by providing access to the OSI environment. The applications layer provides distributed information services and controls the sequence of activities within an application and also the sequence of events between the computer application and the user of another application. The ap- plication layer (shown in Figure 14) communicates directly with the user’s application program. User application processes require application layer service elements to access the net- working environment. There are two types of service elements: CASEs (common applica- tion service elements), which are generally useful to a variety of application processes and SASEs (specific application service elements), which generally satisfy particular needs of application processes. CASE examples include association control that establishes, main- tains, and terminates connections with a peer application entity and commitment, concur- rence, and recovery that ensure the integrity of distributed transactions. SASE examples in- volve the TCP/IP protocol stack and include FTP (file transfer protocol), SNMP (simple network management protocol), Telnet (virtual terminal protocol), and SMTP (simple mail transfer protocol). Introduction to Data Communications and Networking 128
- 134. 7 DATA COMMUNICATIONS CIRCUITS The underlying purpose of a data communications circuit is to provide a transmission path between locations and to transfer digital information from one station to another us- ing electronic circuits. A station is simply an endpoint where subscribers gain access to the circuit. A station is sometimes called a node, which is the location of computers, computer terminals, workstations, and other digital computing equipment. There are al- most as many types of data communications circuits as there are types of data commu- nications equipment. Data communications circuits utilize electronic communications equipment and fa- cilities to interconnect digital computer equipment. Communications facilities are physical means of interconnecting stations within a data communications system and can include virtually any type of physical transmission media or wireless radio system in existence. Communications facilities are provided to data communications users through public tele- phone networks (PTN), public data networks (PDN), and a multitude of private data com- munications systems. Figure 15 shows a simplified block diagram of a two-station data communications circuit. The fundamental components of the circuit are source of digital information, trans- mitter, transmission medium, receiver, and destination for the digital information.Although the figure shows transmission in only one direction, bidirectional transmission is possible by providing a duplicate set of circuit components in the opposite direction. Source. The information source generates data and could be a mainframe computer, personal computer, workstation, or virtually any other piece of digital equipment. The source equipment provides a means for humans to enter data into the system. Transmitter. Source data is seldom in a form suitable to propagate through the trans- mission medium. For example, digital signals (pulses) cannot be propagated through a wireless radio system without being converted to analog first. The transmitter en- codes the source information and converts it to a different form, allowing it to be more efficiently propagated through the transmission medium. In essence, the transmitter acts as an interface between the source equipment and the transmission medium. Transmission medium. The transmission medium carries the encoded signals from the transmitter to the receiver. There are many different types of transmission media, such as free-space radio transmission (including all forms of wireless transmission, such as terrestrial microwave, satellite radio, and cellular telephone) and physical fa- cilities, such as metallic and optical fiber cables. Very often, the transmission path is comprised of several different types of transmission facilities. Receiver. The receiver converts the encoded signals received from the transmission medium back to their original form (i.e., decodes them) or whatever form is used in the destination equipment. The receiver acts as an interface between the transmission medium and the destination equipment. Destination. Like the source, the destination could be a mainframe computer, per- sonal computer, workstation, or virtually any other piece of digital equipment. Introduction to Data Communications and Networking Receiver Digital information destination Transmission medium Transmitter Digital information source FIGURE 15 Simplified block diagram of a two-station data communications circuit 129
- 135. MSB A3 0 Transmitted data LSB Station A A2 1 A1 1 A0 A3 Station B A2 A1 A0 TC 0 Clock (a) Clock 0 1 1 0 Transmitted data Station A output Station B input TC Clock Clock TC TC TC (b) FIGURE 16 Data transmission: (a) parallel; (b) serial 8 SERIAL AND PARALLEL DATA TRANSMISSION Binary information can be transmitted either in parallel or serially. Figure 16a shows how the binary code 0110 is transmitted from station A to station B in parallel. As the figure shows, each bit position (A0 toA3) has its own transmission line. Consequently, all four bits can be transmitted simultaneously during the time of a single clock pulse (TC). This type of transmission is called parallel by bit or serial by character. Figure 16b shows the same binary code transmitted serially. As the figure shows, there is a single transmission line and, thus, only one bit can be transmitted at a time. Con- sequently, it requires four clock pulses (4TC) to transmit the entire four-bit code. This type of transmission is called serial by bit. Obviously, the principal trade-off between parallel and serial data transmission is speed versus simplicity. Data transmission can be accomplished much more rapidly using parallel transmission; however, parallel transmission requires more data lines. As a general rule, parallel transmission is used for short-distance data communications and within a computer, and serial transmission is used for long-distance data communications. 9 DATA COMMUNICATIONS CIRCUIT ARRANGEMENTS Data communications circuits can be configured in a multitude of arrangements depending on the specifics of the circuit, such as how many stations are on the circuit, type of transmis- sion facility, distance between stations, and how many users are at each station. A data com- munications circuit can be described in terms of circuit configuration and transmission mode. 9-1 Circuit Configurations Data communications networks can be generally categorized as either two point or multi- point.A two-point configuration involves only two locations or stations, whereas a multipoint configuration involves three or more stations. Regardless of the configuration, each station can have one or more computers, computer terminals, or workstations. A two-point circuit involves the transfer of digital information between a mainframe computer and a personal computer, two mainframe computers, two personal computers, or two data communications networks. A multipoint network is generally used to interconnect a single mainframe com- puter (host) to many personal computers or to interconnect many personal computers. 9-2 Transmission Modes Essentially, there are four modes of transmission for data communications circuits: simplex, half duplex, full duplex, and full/full duplex. Introduction to Data Communications and Networking 130
- 136. 9-2-1 Simplex. In the simplex (SX) mode, data transmission is unidirectional; in- formation can be sent in only one direction. Simplex lines are also called receive-only, transmit-only, or one-way-only lines. Commercial radio broadcasting is an example of sim- plex transmission, as information is propagated in only one direction—from the broadcast- ing station to the listener. 9-2-2 Half duplex. In the half-duplex (HDX) mode, data transmission is possible in both directions but not at the same time. Half-duplex communications lines are also called two-way-alternate or either-way lines. Citizens band (CB) radio is an example of half-duplex transmission because to send a message, the push-to-talk (PTT) switch must be depressed, which turns on the transmitter and shuts off the receiver. To receive a message, the PTT switch must be off, which shuts off the transmitter and turns on the receiver. 9-2-3 Full duplex. In the full-duplex (FDX) mode, transmissions are possible in both directions simultaneously, but they must be between the same two stations. Full-du- plex lines are also called two-way simultaneous, duplex, or both-way lines. A local tele- phone call is an example of full-duplex transmission. Although it is unlikely that both par- ties would be talking at the same time, they could if they wanted to. 9-2-4 Full/full duplex. In the full/full duplex (F/FDX) mode, transmission is pos- sible in both directions at the same time but not between the same two stations (i.e., one sta- tion is transmitting to a second station and receiving from a third station at the same time). Full/full duplex is possible only on multipoint circuits. The U.S. postal system is an exam- ple of full/full duplex transmission because a person can send a letter to one address and re- ceive a letter from another address at the same time. 10 DATA COMMUNICATIONS NETWORKS Any group of computers connected together can be called a data communications network, and the process of sharing resources between computers over a data communications net- work is called networking. In its simplest form, networking is two or more computers con- nected together through a common transmission medium for the purpose of sharing data.The concept of networking began when someone determined that there was a need to share soft- ware and data resources and that there was a better way to do it than storing data on a disk and literally running from one computer to another. By the way, this manual technique of mov- ing data on disks is sometimes referred to as sneaker net. The most important considerations of a data communications network are performance, transmission rate, reliability, and security. Applications running on modern computer networks vary greatly from company to company.A network must be designed with the intended application in mind.A general cat- egorization of networking applications is listed in Table 1. The specific application affects how well a network will perform. Each network has a finite capacity. Therefore, network designers and engineers must be aware of the type and frequency of information traffic on the network. Introduction to Data Communications and Networking Table 1 Networking Applications Application Examples Standard office applications E-mail, file transfers, and printing High-end office applications Video imaging, computer-aided drafting, computer-aided design, and soft- ware development Manufacturing automation Process and numerical control Mainframe connectivity Personal computers, workstations, and terminal support Multimedia applications Live interactive video 131
- 137. Local area networks Wide area networks Metropolitan area networks Global area networks End stations Applications Networks FIGURE 17 Basic network components There are many factors involved when designing a computer network, including the following: 1. Network goals as defined by organizational management 2. Network security 3. Network uptime requirements 4. Network response-time requirements 5. Network and resource costs The primary balancing act in computer networking is speed versus reliability. Too often, network performance is severely degraded by using error checking procedures, data en- cryption, and handshaking (acknowledgments). However, these features are often required and are incorporated into protocols. Some networking protocols are very reliable but require a significant amount of over- head to provide the desired high level of service. These protocols are examples of connection- oriented protocols. Other protocols are designed with speed as the primary parameter and, therefore, forgo some of the reliability features of the connection-oriented protocols. These quick protocols are examples of connectionless protocols. 10-1 Network Components, Functions, and Features Computer networks are like snowflakes—no two are the same. The basic components of computer networks are shown in Figure 17. All computer networks include some combi- nation of the following: end stations, applications, and a network that will support the data traffic between the end stations. A computer network designed three years ago to support the basic networking applications of the time may have a difficult time supporting recently Introduction to Data Communications and Networking 132
- 138. User computer File server File request Copy of requested file FIGURE 18 File server operation developed high-end applications, such as medical imaging and live video teleconferencing. Network designers, administrators, and managers must understand and monitor the most recent types and frequency of networked applications. Computer networks all share common devices, functions, and features, including servers, clients, transmission media, shared data, shared printers and other peripherals, hardware and software resources, network interface card (NIC), local operating system (LOS), and the network operating system (NOS). 10-1-1 Servers. Servers are computers that hold shared files, programs, and the net- work operating system. Servers provide access to network resources to all the users of the network. There are many different kinds of servers, and one server can provide several func- tions. For example, there are file servers, print servers, mail servers, communications servers, database servers, directory/security servers, fax servers, and Web servers, to name a few. Figure 18 shows the operation of a file server. A user (client) requests a file from the file server. The file server sends a copy of the file to the requesting user. File servers allow users to access and manipulate disk resources stored on other computers. An example of a file server application is when two or more users edit a shared spreadsheet file that is stored on a server. File servers have the following characteristics: 1. File servers are loaded with files, accounts, and a record of the access rights of users or groups of users on the network. 2. The server provides a shareable virtual disk to the users (clients). 3. File mapping schemes are implemented to provide the virtualness of the files (i.e., the files are made to look like they are on the user’s computer). 4. Security systems are installed and configured to provide the server with the re- quired security and protection for the files. 5. Redirector or shell software programs located on the users’ computers transpar- ently activate the client’s software on the file server. 10-1-2 Clients. Clients are computers that access and use the network and shared network resources. Client computers are basically the customers (users) of the network, as they request and receive services from the servers. 10-1-3 Transmission media. Transmission media are the facilities used to inter- connect computers in a network, such as twisted-pair wire, coaxial cable, and optical fiber cable. Transmission media are sometimes called channels, links, or lines. 10-1-4 Shared data. Shared data are data that file servers provide to clients, such as data files, printer access programs, and e-mail. 10-1-5 Shared printers and other peripherals. Shared printers and peripherals are hardware resources provided to the users of the network by servers. Resources provided include data files, printers, software, or any other items used by clients on the network. Introduction to Data Communications and Networking 133
- 139. Computer 04 MAC (media access control) address (six bytes – 12 hex characters – 48 bits) NIC Card 60 8C 49 F2 3B FIGURE 19 Network interface card (NIC) 10-1-6 Network interface card. Each computer in a network has a special expan- sion card called a network interface card (NIC). The NIC prepares (formats) and sends data, receives data, and controls data flow between the computer and the network. On the trans- mit side, the NIC passes frames of data on to the physical layer, which transmits the data to the physical link. On the receive side, the NIC processes bits received from the physical layer and processes the message based on its contents. A network interface card is shown in Figure 19. Characteristics of NICs include the following: 1. The NIC constructs, transmits, receives, and processes data to and from a PC and the connected network. 2. Each device connected to a network must have a NIC installed. 3. A NIC is generally installed in a computer as a daughterboard, although some com- puter manufacturers incorporate the NIC into the motherboard during manufacturing. 4. Each NIC has a unique six-byte media access control (MAC) address, which is typically permanently burned into the NIC when it is manufactured. The MAC ad- dress is sometimes called the physical, hardware, node, Ethernet, or LAN address. 5. The NIC must be compatible with the network (i.e., Ethernet—10baseT or token ring) to operate properly. 6. NICs manufactured by different vendors vary in speed, complexity, manageabil- ity, and cost. 7. The NIC requires drivers to operate on the network. 10-1-7 Local operating system. A local operating system (LOS) allows per- sonal computers to access files, print to a local printer, and have and use one or more disk and CD drives that are located on the computer. Examples of LOSs are MS-DOS, PC-DOS, Unix, Macintosh, OS/2, Windows 3.11, Windows 95, Windows 98, Windows 2000, and Linux. Figure 20 illustrates the relationship between a personal computer and its LOS. 10-1-8 Network operating system. The network operating system (NOS) is a pro- gramthatrunsoncomputersandserversthatallowsthecomputerstocommunicateoveranet- work. The NOS provides services to clients such as log-in features, password authentication, Introduction to Data Communications and Networking 134
- 140. Personal computer Windows UNIX MS-DOS Macintosh FIGURE 20 Local operating system (LOS) Server Client Client NOS NOS NOS FIGURE 21 Network operating system (NOS) printer access, network administration functions, and data file sharing. Some of the more pop- ular network operating systems are Unix, Novell NetWare,AppleShare, Macintosh System 7, IBM LAN Server, Compaq OpenVMS, and MicrosoftWindows NT Server.The NOS is soft- warethatmakescommunicationsoveranetworkmoremanageable.Therelationshipbetween clients, servers, and the NOS is shown in Figure 21, and the layout of a local network operat- ing system is depicted in Figure 22. Characteristics of NOSs include the following: 1. A NOS allows users of a network to interface with the network transparently. 2. A NOS commonly offers the following services: file service, print service, mail ser- vice, communications service, database service, and directory and security services. 3. The NOS determines whether data are intended for the user’s computer or whether the data needs to be redirected out onto the network. 4. The NOS implements client software for the user, which allows them to access servers on the network. 10-2 Network Models Computer networks can be represented with two basic network models: peer-to-peer client/server and dedicated client/server. The client/server method specifies the way in which two computers can communicate with software over a network.Although clients and servers are generally shown as separate units, they are often active in a single computer but not at the same time.With the client/server concept, a computer acting as a client initiates a software re- quest from another computer acting as a server. The server computer responds and attempts Introduction to Data Communications and Networking 135
- 141. Introduction to Data Communications and Networking User 1 Database server File server Communications server Mail server Print server Printer User 5 User 2 User 4 User 3 NOS To other networks and servers Hub FIGURE 22 Network layout using a network operating system (NOS) Hub Client/server 1 Client/server 2 FIGURE 23 Client/server concept to satisfy the request from the client. The server computer might then act as a client and re- quest services from another computer. The client/server concept is illustrated in Figure 23. 10-2-1 Peer-to-peer client/server network. A peer-to-peer client/server network is one in which all computers share their resources, such as hard drives, printers, and so on, with all the other computers on the network. Therefore, the peer-to-peer operating sys- tem divides its time between servicing the computer on which it is loaded and servicing 136
- 142. Introduction to Data Communications and Networking Hub Client/server 1 Client/server 2 Client/server 4 Client/server 3 FIGURE 24 Peer-to-peer client/server network requests from other computers. In a peer-to-peer network (sometimes called a workgroup), there are no dedicated servers or hierarchy among the computers. Figure 24 shows a peer-to-peer client/server network with four clients/servers (users) connected together through a hub. All computers are equal, hence the name peer. Each computer in the network can function as a client and/or a server, and no sin- gle computer holds the network operating system or shared files. Also, no one com- puter is assigned network administrative tasks. The users at each computer determine which data on their computer are shared with the other computers on the network. In- dividual users are also responsible for installing and upgrading the software on their computer. Because there is no central controlling computer, a peer-to-peer network is an appro- priate choice when there are fewer than 10 users on the network, when all computers are lo- cated in the same general area, when security is not an issue, or when there is limited growth projected for the network in the immediate future. Peer-to-peer computer networks should be small for the following reasons: 1. When operating in the server role, the operating system is not optimized to effi- ciently handle multiple simultaneous requests. 2. The end user’s performance as a client would be degraded. 3. Administrative issues such as security, data backups, and data ownership may be compromised in a large peer-to-peer network. 10-2-2 Dedicated client/server network. In a dedicated client/server network, one computer is designated the server, and the rest of the computers are clients. As the network grows, additional computers can be designated servers. Generally, the designated servers function only as servers and are not used as a client or workstation. The servers store all the network’s shared files and applications programs, such as word processor documents, com- pilers, database applications, spreadsheets, and the network operating system. Client com- puters can access the servers and have shared files transferred to them over the transmission medium. Figure 25 shows a dedicated client/server-based network with three servers and three clients (users). Each client can access the resources on any of the servers and also the re- sources on other client computers. The dedicated client/server-based network is probably 137
- 143. Introduction to Data Communications and Networking Hub Client 1 Client 2 Client 3 Dedicated file server Dedicated print server Dedicated mail server Printer FIGURE 25 Dedicated client/server network the most commonly used computer networking model. There can be a separate dedicated server for each function (i.e., file server, print server, mail server, etc.) or one single general- purpose server responsible for all services. In some client/server networks, client computers submit jobs to one of the servers. The server runs the software and completes the job and then sends the results back to the client computer. In this type of client/server network, less information propagates through the network than with the file server configuration because only data and not applications programs are transferred between computers. Ingeneral,thededicatedclient/servermodelispreferabletothepeer-to-peerclient/server model for general-purpose data networks. The peer-to-peer model client/server model is usu- ally preferable for special purposes, such as a small group of users sharing resources. 10-3 Network Topologies Network topology describes the layout or appearance of a network—that is, how the com- puters, cables, and other components within a data communications network are intercon- nected, both physically and logically. The physical topology describes how the network is actually laid out, and the logical topology describes how data actually flow through the network. In a data communications network, two or more stations connect to a link, and one or more links form a topology. Topology is a major consideration for capacity, cost, and reli- ability when designing a data communications network. The most basic topologies are point to point and multipoint. A point-to-point topology is used in data communications networks that transfer high-speed digital information between only two stations. Very of- ten, point-to-point data circuits involve communications between a mainframe computer and another mainframe computer or some other type of high-capacity digital device. A two- point circuit is shown in Figure 26a. A multipoint topology connects three or more stations through a single transmission medium. Examples of multipoint topologies are star, bus, ring, mesh, and hybrid. 138
- 144. Bus Ring (f) (d) (c) Bus (b) (a) (e) Hub FIGURE 26 Network topologies: (a) point-to-point; (b) star; (c) bus; (d) ring; (e) mesh; (f) hybrid 139
- 145. Introduction to Data Communications and Networking 10-3-1 Star topology. A star topology is a multipoint data communications net- work where remote stations are connected by cable segments directly to a centrally located computer called a hub, which acts like a multipoint connector (see Figure 26b). In essence, a star topology is simply a multipoint circuit comprised of many two-point circuits where each remote station communicates directly with a centrally located computer. With a star topology, remote stations cannot communicate directly with one another, so they must re- lay information through the hub. Hubs also have store-and-forward capabilities, enabling them to handle more than one message at a time. 10-3-2 Bus topology. A bus topology is a multipoint data communications circuit that makes it relatively simple to control data flow between and among the computers be- cause this configuration allows all stations to receive every transmission over the network. With a bus topology, all the remote stations are physically or logically connected to a sin- gle transmission line called a bus. The bus topology is the simplest and most common method of interconnecting computers. The two ends of the transmission line never touch to form a complete loop. A bus topology is sometimes called multidrop or linear bus, and all stations share a common transmission medium. Data networks using the bus topology generally involve one centrally located host computer that controls data flow to and from the other stations. The bus topology is sometimes called a horizontal bus and is shown in Figure 26c. 10-3-3 Ring topology. A ring topology is a multipoint data communications net- work where all stations are interconnected in tandem (series) to form a closed loop or cir- cle. A ring topology is sometimes called a loop. Each station in the loop is joined by point- to-point links to two other stations (the transmitter of one and the receiver of the other) (see Figure 26d). Transmissions are unidirectional and must propagate through all the stations in the loop. Each computer acts like a repeater in that it receives signals from down-line computers then retransmits them to up-line computers. The ring topology is similar to the bus and star topologies, as it generally involves one centrally located host computer that controls data flow to and from the other stations. 10-3-4 Mesh topology. In a mesh topology, every station has a direct two-point communications link to every other station on the circuit as shown in Figure 26e. The mesh topology is sometimes called fully connected. A disadvantage of a mesh topol- ogy is a fully connected circuit requires n(n 1)/2 physical transmission paths to in- terconnect n stations and each station must have n 1 input/output ports. Advantages of a mesh topology are reduced traffic problems, increased reliability, and enhanced security. 10-3-5 Hybrid topology. A hybrid topology is simply combining two or more of the traditional topologies to form a larger, more complex topology. Hybrid topologies are sometimes called mixed topologies. An example of a hybrid topology is the bus star topol- ogy shown in Figure 26f. Other hybrid configurations include the star ring, bus ring, and virtually every other combination you can think of. 10-4 Network Classifications Networks are generally classified by size, which includes geographic area, distance between stations, number of computers, transmission speed (bps), transmission me- dia, and the network’s physical architecture. The four primary classifications of net- works are local area networks (LANs), metropolitan area networks (MANs), wide 140
- 146. Introduction to Data Communications and Networking Table 2 Primary Network Types Network Type Characteristics LAN (local area network) Interconnects computer users within a department, company, or group MAN (metropolitan area network) Interconnects computers in and around a large city WAN (wide area network) Interconnects computers in and around an entire country GAN (global area network) Interconnects computers from around the entire globe Building backbone Interconnects LANs within a building Campus backbone Interconnects building LANs Enterprise network Interconnects many or all of the above PAN (personal area network) Interconnects memory cards carried by people and in computers that are in close proximity to each other PAN (power line area network, Virtually no limit on how many computers it can interconnect sometimes called PLAN) and covers an area limited only by the availability of power distribution lines area networks (WANs), and global area networks (GANs). In addition, there are three primary types of interconnecting networks: building backbone, campus backbone, and enterprise network. Two promising computer networks of the future share the same acronym: the PAN (personal area network) and PAN (power line area network, sometimes called PLAN). The idea behind a personal area network is to allow people to transfer data through the human body simply by touching each other. Power line area networks use existing ac distribution networks to carry data wherever power lines go, which is virtually everywhere. When two or more networks are connected together, they constitute an internetwork or internet. An internet (lowercase i) is sometimes confused with the Internet (uppercase I). The term internet is a generic term that simply means to interconnect two or more net- works, whereas Internet is the name of a specific worldwide data communications net- work. Table 2 summarizes the characteristics of the primary types of networks, and Figure 27 illustrates the geographic relationship among computers and the different types of net- works. 10-4-1 Local area network. Local area networks (LANs) are typically privately owned data communications networks in which 10 to 40 computer users share data re- sources with one or more file servers. LANs use a network operating system to provide two-way communications at bit rates typically in the range of 10 Mbps to 100 Mbps and higher between a large variety of data communications equipment within a relatively small geographical area, such as in the same room, building, or building complex (see Figure 28). A LAN can be as simple as two personal computers and a printer or could contain dozens of computers, workstations, and peripheral devices. Most LANs link equipment that are within a few miles of each other or closer. Because the size of most LANs is limited, the longest (or worst-case) transmission time is bounded and known by everyone using the network. Therefore, LANs can utilize configurations that otherwise would not be possible. LANs were designed for sharing resources between a wide range of digital equip- ment, including personal computers, workstations, and printers. The resources shared can be software as well as hardware. Most LANs are owned by the company or organization 141
- 147. Introduction to Data Communications and Networking Metropolitan area network Multiple buildings or entire city Wide area network Entire country Global area network Entire world Personal area network Between people and computers Local area network Single building FIGURE 27 Computer network types that uses it and have a connection to a building backbone for access to other departmental LANs, MANs, WANs, and GANs. 10-4-2 Metropolitan area network. A metropolitan area network (MAN) is a high-speed network similar to a LAN except MANs are designed to encompass larger areas, usually that of an entire city (see Figure 29). Most MANs support the trans- mission of both data and voice and in some cases video. MANs typically operate at 142
- 148. Introduction to Data Communications and Networking CD-ROM/WORM FAX machine File/application/ print server Hub/repeater To building backbone Router or switch Patch panel Wall jack LAN Scanner Laptop PC Workstation NOS client software NOS client software NOS server FIGURE 28 Local area network (LAN) layout speeds of 1.5 Mbps to 10 Mbps and range from five miles to a few hundred miles in length. A MAN generally uses only one or two transmission cables and requires no switches. A MAN could be a single network, such as a cable television distribution net- work, or it could be a means of interconnecting two or more LANs into a single, larger network, enabling data resources to be shared LAN to LAN as well as from station to station or computer to computer. Large companies often use MANS to interconnect all their LANs. A MAN can be owned and operated entirely by a single, private company, or it could lease services and facilities on a monthly basis from the local cable or telephone company. Switched Multimegabit Data Services (SMDS) is an example of a service of- fered by local telephone companies for handling high-speed data communications for MANs. Other examples of MANs are FDDI (fiber distributed data interface) and ATM (asynchronous transfer mode). 10-4-3 Wide area network. Wide area networks (WANs) are the oldest type of data communications network that provide relatively slow-speed, long-distance transmission of data, voice, and video information over relatively large and widely dispersed geographical areas, such as a country or an entire continent (see Figure 30). WANs typically interconnect cities and states. WANs typically operate at bit rates from 1.5 Mbps to 2.4 Gbps and cover a distance of 100 to 1000 miles. WANs may utilize both public and private communications systems to provide serv- ice over an area that is virtually unlimited; however, WANs are generally obtained through service providers and normally come in the form of leased-line or circuit-switching tech- nology. Often WANs interconnect routers in different locations. Examples of WANs are 143
- 149. Introduction to Data Communications and Networking FIGURE 29 Metropolitan area network (MAN) ISDN (integrated services digital network), T1 and T3 digital carrier systems, frame relay, X.25, ATM, and using data modems over standard telephone lines. 10-4-4 Global area network. Global area networks (GANs) provide connects be- tween countries around the entire globe (see Figure 31). The Internet is a good example of a GAN, as it is essentially a network comprised of other networks that interconnects virtu- ally every country in the world. GANs operate from 1.5 Mbps to 100 Gbps and cover thou- sands of miles 10-4-5 Building backbone. A building backbone is a network connection that nor- mally carries traffic between departmental LANs within a single company.A building back- bone generally consists of a switch or a router (see Figure 32) that can provide connectiv- ity to other networks, such as campus backbones, enterprise backbones, MANs, WANs, or GANs. Phoenix metropolitan area Manufacturing facility Research facility Shipping facility Headquarters building Service provider MAN 144
- 150. Introduction to Data Communications and Networking Service provider WAN Seatle, WA Tempe, AZ San Diego, CA Oriskany, NY Router Router Router Router Router Miami, FL FIGURE 30 Wide area network (WAN) 10-4-6 Campus backbone. A campus backbone is a network connection used to carry traffic to and from LANs located in various buildings on campus (see Figure 33). A campus backbone is designed for sites that have a group of buildings at a single location, such as corporate headquarters, universities, airports, and research parks. A campus backbone normally uses optical fiber cables for the transmission media be- tween buildings. The optical fiber cable is used to connect interconnecting devices, such as Loa Angeles, CA – USA Sidney, Australia Rome, Italy London, England GAN FIGURE 31 Global area network (GAN) 145
- 151. Introduction to Data Communications and Networking Hub Building backbone optical fiber cables PC Wall jacks LAN Hub Workstation LAN Switch Router To WAN, MAN, or campus Patch cables Patch panels FIGURE 32 Building backbone Router/switch Building 1 Router/switch Router/switch LAN LANs Fiber cables LANs Router/switch Building 2 Building 3 WAN WAN FIGURE 33 Campus backbone bridges, routers, and switches. Campus backbones must operate at relatively high trans- mission rates to handle the large volumes of traffic between sites. 10-4-7 Enterprise networks. An enterprise network includes some or all of the previ- ously mentioned networks and components connected in a cohesive and manageable fashion. 146
- 152. Introduction to Data Communications and Networking 11 ALTERNATE PROTOCOL SUITES The functional layers of the OSI seven-layer protocol hierarchy do not line up well with certain data communications applications, such as the Internet. Because of this, there are several other protocolsthatseewidespreaduse,suchasTCP/IPandtheCiscothree-layerhierarchicalmodel. 11-1 TCP/IP Protocol Suite The TCP/IP protocol suite (transmission control protocol/Internet protocol) was actu- ally developed by the Department of Defense before the inception of the seven-layer OSI model. TCP/IP is comprised of several interactive modules that provide specific functionality without necessarily operating independent of one another. The OSI seven- layer model specifies exactly which function each layer performs, whereas TCP/IP is comprised of several relatively independent protocols that can be combined in many ways, depending on system needs. The term hierarchical simply means that the upper- level protocols are supported by one or more lower-level protocols. Depending on whose definition you use, TCP/IP is a hierarchical protocol comprised of either three or four layers. The three-layer version of TCP/IP contains the network, transport, and application layers that reside above two lower-layer protocols that are not specified by TCP/IP (the physical and data link layers). The network layer of TCP/IP provides internetworking func- tions similar to those provided by the network layer of the OSI network model. The net- work layer is sometimes called the internetwork layer or internet layer. The transport layer of TCP/IP contains two protocols: TCP (transmission control pro- tocol) and UDP (user datagram protocol). TCP functions go beyond those specified by the transport layer of the OSI model, as they define several tasks defined for the session layer. In essence, TCP allows two application layers to communicate with each other. The applications layer of TCP/IP contains several other protocols that users and pro- grams utilize to perform the functions of the three uppermost layers of the OSI hierarchy (i.e., the applications, presentation, and session layers). The four-layer version of TCP/IP specifies the network access, Internet, host-to-host, and process layers: Network access layer. Provides a means of physically delivering data packets using frames or cells Internet layer. Contains information that pertains to how data can be routed through the network Host-to-host layer. Services the process and Internet layers to handle the reliability and session aspects of data transmission Process layer. Provides applications support TCP/IP is probably the dominant communications protocol in use today. It provides a common denominator, allowing many different types of devices to communicate over a network or system of networks while supporting a wide variety of applications. 11-2 Cisco Three-Layer Model Cisco defines a three-layer logical hierarchy that specifies where things belong, how they fit together, and what functions go where. The three layers are the core, distribution, and access: Core layer. The core layer is literally the core of the network, as it resides at the top of the hierarchy and is responsible for transporting large amounts of data traffic reli- ably and quickly. The only purpose of the core layer is to switch traffic as quickly as possible. 147
- 153. Introduction to Data Communications and Networking Distribution layer. The distribution layer is sometimes called the workgroup layer. The distribution layer is the communications point between the access and the core layers that provides routing, filtering, WAN access, and how many data packets are allowed to access the core layer. The distribution layer determines the fastest way to handle service requests, for example, the fastest way to forward a file request to a server. Several functions are performed at the distribution level: 1. Implementation of tools such as access lists, packet filtering, and queuing 2. Implementation of security and network policies, including firewalls and address translation 3. Redistribution between routing protocols 4. Routing between virtual LANS and other workgroup support functions 5. Define broadcast and multicast domains Access layer. The access layer controls workgroup and individual user access to inter- networking resources, most of which are available locally. The access layer is some- times called the desktop layer. Several functions are performed at the access layer level: 1. Access control 2. Creation of separate collision domains (segmentation) 3. Workgroup connectivity into the distribution layer QUESTIONS 1. Define the following terms: data, information, and data communications network. 2. What was the first data communications system that used binary-coded electrical signals? 3. Discuss the relationship between network architecture and protocol. 4. Briefly describe broadcast and point-to-point computer networks. 5. Define the following terms: protocol, connection-oriented protocols, connectionless protocols, and protocol stacks. 6. What is the difference between syntax and semantics? 7. What are data communications standards, and why are they needed? 8. Name and briefly describe the differences between the two kinds of data communications stan- dards. 9. List and describe the eight primary standards organizations for data communications. 10. Define the open systems interconnection. 11. Briefly describe the seven layers of the OSI protocol hierarchy. 12. List and briefly describe the basic functions of the five components of a data communications cir- cuit. 13. Briefly describe the differences between serial and parallel data transmission. 14. What are the two basic kinds of data communications circuit configurations? 15. List and briefly describe the four transmission modes. 16. List and describe the functions of the most common components of a computer network. 17. What are the differences between servers and clients on a data communications network? 18. Describe a peer-to-peer data communications network. 19. What are the differences between peer-to-peer client/server networks and dedicated client/server networks? 20. What is a data communications network topology? 21. List and briefly describe the five basic data communications network topologies. 22. List and briefly describe the major network classifications. 23. Briefly describe the TCP/IP protocol model. 24. Briefly describe the Cisco three-layer protocol model. 148
- 154. Fundamental Concepts of Data Communications CHAPTER OUTLINE 1 Introduction 8 Data Communications Hardware 2 Data Communications Codes 9 Data Communications Circuits 3 Bar Codes 10 Line Control Unit 4 Error Control 11 Serial Interfaces 5 Error Detection 12 Data Communications Modems 6 Error Correction 13 ITU-T Modem Recommendations 7 Character Synchronization OBJECTIVES ■ Define data communication code ■ Describe the following data communications codes: Baudot, ASCII, and EBCDIC ■ Explain bar code formats ■ Define error control, error detection, and error correction ■ Describe the following error-detection mechanisms: redundancy, checksum, LRC, VRC, and CRC ■ Describe the following error-correction mechanisms: FEC, ARQ, and Hamming code ■ Describe character synchronization and explain the differences between asynchronous and synchronous data formats ■ Define the term data communications hardware ■ Describe data terminal equipment ■ Describe data communications equipment ■ List and describe the seven components that make up a two-point data communications circuit ■ Describe the terms line control unit and front-end processor and explain the differences between the two ■ Describe the basic operation of a UART and outline the differences between UARTs, USRTs, and USARTs ■ Describe the functions of a serial interface ■ Explain the physical, electrical, and functional characteristics of the RS-232 serial interface ■ Compare and contrast the RS-232, RS-449, and RS-530 serial interfaces From Chapter 4 of Advanced Electronic Communications Systems, Sixth Edition. Wayne Tomasi. Copyright © 2004 by Pearson Education, Inc. Published by Pearson Prentice Hall. All rights reserved. 149
- 155. ■ Describe data communications modems ■ Explain the block diagram of a modem ■ Explain what is meant by Bell System–compatible modems ■ Describe modem synchronization and modem equalization ■ Describe the ITU-T modem recommendations 1 INTRODUCTION To understand how a data communications network works as an entity, it is necessary first to understand the fundamental concepts and components that make up the network. The fundamental concepts of data communications include data communications code, error control (error detection and correction), and character synchronization, and fundamental hardware includes various pieces of computer and networking equipment, such as line con- trol units, serial interfaces, and data communications modems. 2 DATA COMMUNICATIONS CODES Data communications codes are often used to represent characters and symbols, such as let- ters, digits, and punctuation marks. Therefore, data communications codes are called character codes, character sets, symbol codes, or character languages. 2-1 Baudot Code The Baudot code (sometimes called the Telex code) was the first fixed-length character code developed for machines rather than for people. A French postal engineer named Thomas Murray developed the Baudot code in 1875 and named the code after Emile Bau- dot, an early pioneer in telegraph printing. The Baudot code (pronounced baw-dough) is a fixed-length source code (sometimes called a fixed-length block code). With fixed-length source codes, all characters are represented in binary and have the same number of symbols (bits). The Baudot code is a five-bit character code that was used primarily for low-speed teletype equipment, such as the TWX/Telex system and radio teletype (RTTY). The latest version of the Baudot code is recommended by the CCITT as the International Alphabet No. 2 and is shown in Table 1. 2-2 ASCII Code In 1963, in an effort to standardize data communications codes, the United States adopted the Bell System model 33 teletype code as the United States of America Standard Code for Information Exchange (USASCII), better known as ASCII-63. Since its adoption, ASCII (pronounced as-key) has progressed through the 1965, 1967, and 1977 versions, with the 1977 version being recommended by the ITU as InternationalAlphabet No. 5, in the United States as ANSI standard X3.4-1986 (R1997), and by the International Standards Organiza- tion as ISO-14962 (1997). ASCII is the standard character set for source coding the alphanumeric character set that humans understand but computers do not (computers only understand 1s and 0s). ASCII is a seven-bit fixed-length character set. With the ASCII code, the least-significant bit (LSB) is designated b0 and the most-significant bit (MSB) is designated b7 as shown here: b7 b6 b5 b4 b3 b2 b1 b0 MSB LSB Direction of propagation The terms least and most significant are somewhat of a misnomer because character codes do not represent weighted binary numbers and, therefore, all bits are equally sig- Fundamental Concepts of Data Communications 150
- 156. nificant. Bit b7 is not part of the ASCII code but is generally reserved for an error detec- tion bit called the parity bit, which is explained later in this chapter. With character codes, it is more meaningful to refer to bits by their order than by their position; b0 is the zero- order bit, b1 the first-order bit, b7 the seventh-order bit, and so on. However, with serial data transmission, the bit transmitted first is generally called the LSB. With ASCII, the low-order bit (b0) is transmitted first. ASCII is probably the code most often used in data communications networks today. The 1977 version of the ASCII code with odd parity is shown in Table 2 (note that the parity bit is not included in the hex code). 2-3 EBCDIC Code The extended binary-coded decimal interchange code (EBCDIC) is an eight-bit fixed- length character set developed in 1962 by the International Business Machines Corporation (IBM). EBCDIC is used almost exclusively with IBM mainframe computers and peripheral equipment. With eight bits, 28 , or 256, codes are possible, although only 139 of the 256 codes are actually assigned characters. Unspecified codes can be assigned to specialized characters and functions. The name binary coded decimal was selected because the second hex character for all letter and digit codes contains only the hex values from 0 to 9, which have the same binary sequence as BCD codes. The EBCDIC code is shown in Table 3. Fundamental Concepts of Data Communications Table 1 Baudot Code Bit Letter Figure Bit: 4 3 2 1 0 A — 1 1 0 0 0 B ? 1 0 0 1 1 C : 0 1 1 1 0 D $ 1 0 0 1 0 E 3 1 0 0 0 0 F ! 1 0 1 1 0 G 0 1 0 1 1 H # 0 0 1 0 1 I 8 0 1 1 0 0 J ' 1 1 0 1 0 K ( 1 1 1 1 0 L ) 0 1 0 0 1 M . 0 0 1 1 1 N , 0 0 1 1 0 O 9 0 0 0 1 1 P 0 0 1 1 0 1 Q 1 1 1 1 0 1 R 4 0 1 0 1 0 S bel 1 0 1 0 0 T 5 0 0 0 0 1 U 7 1 1 1 0 0 V ; 0 1 1 1 1 W 2 1 1 0 0 1 X / 1 0 1 1 1 Y 6 1 0 1 0 1 Z ″ 1 0 0 0 1 Figure shift 1 1 1 1 1 Letter shift 1 1 0 1 1 Space 0 0 1 0 0 Line feed (LF) 0 1 0 0 0 Blank (null) 0 0 0 0 0 151
- 157. Fundamental Concepts of Data Communications Table 2 ASCII-77: Odd Parity Binary Code Binary Code Bit 7 6 5 4 3 2 1 0 Hex Bit 7 6 5 4 3 2 1 0 Hex NUL 1 0 0 0 0 0 0 0 00 @ 0 1 0 0 0 0 0 0 40 SOH 0 0 0 0 0 0 0 1 01 A 1 1 0 0 0 0 0 1 41 STX 0 0 0 0 0 0 1 0 02 B 1 1 0 0 0 0 1 0 42 ETX 1 0 0 0 0 0 1 1 03 C 0 1 0 0 0 0 1 1 43 EOT 0 0 0 0 0 1 0 0 04 D 1 1 0 0 0 1 0 0 44 ENQ 1 0 0 0 0 1 0 1 05 E 0 1 0 0 0 1 0 1 45 ACK 1 0 0 0 0 1 1 0 06 F 0 1 0 0 0 1 1 0 46 BEL 0 0 0 0 0 1 1 1 07 G 1 1 0 0 0 1 1 1 47 BS 0 0 0 0 1 0 0 0 08 H 1 1 0 0 1 0 0 0 48 HT 1 0 0 0 1 0 0 1 09 I 0 1 0 0 1 0 0 1 49 NL 1 0 0 0 1 0 1 0 0A J 0 1 0 0 1 0 1 0 4A VT 0 0 0 0 1 0 1 1 0B K 1 1 0 0 1 0 1 1 4B FF 1 0 0 0 1 1 0 0 0C L 0 1 0 0 1 1 0 0 4C CR 0 0 0 0 1 1 0 1 0D M 1 1 0 0 1 1 0 1 4D SO 0 0 0 0 1 1 1 0 0E N 1 1 0 0 1 1 1 0 4E SI 1 0 0 0 1 1 1 1 0F O 0 1 0 0 1 1 1 1 4F DLE 0 0 0 1 0 0 0 0 10 P 1 1 0 1 0 0 0 0 50 DC1 0 0 0 1 0 0 0 1 11 Q 0 1 0 1 0 0 0 1 51 DC2 1 0 0 1 0 0 1 0 12 R 0 1 0 1 0 0 1 0 52 DC3 0 0 0 1 0 0 1 1 13 S 1 1 0 1 0 0 1 1 53 DC4 1 0 0 1 0 1 0 0 14 T 0 1 0 1 0 1 0 0 54 NAK 0 0 0 1 0 1 0 1 15 U 1 1 0 1 0 1 0 1 55 SYN 0 0 0 1 0 1 1 0 16 V 1 1 0 1 0 1 1 0 56 ETB 1 0 0 1 0 1 1 1 17 W 0 1 0 1 0 1 1 1 57 CAN 1 0 0 1 1 0 0 0 18 X 0 1 0 1 1 0 0 0 58 EM 0 0 0 1 1 0 0 1 19 Y 1 1 0 1 1 0 0 1 59 SUB 0 0 0 1 1 0 1 0 1A Z 1 1 0 1 1 0 1 0 5A ESC 1 0 0 1 1 0 1 1 1B [ 0 1 0 1 1 0 1 1 5B FS 0 0 0 1 1 1 0 0 1C 1 1 0 1 1 1 0 0 5C GS 1 0 0 1 1 1 0 1 1D ] 0 1 0 1 1 1 0 1 5D RS 1 0 0 1 1 1 1 0 1E 0 1 0 1 1 1 1 0 5E US 0 0 0 1 1 1 1 1 1F - 1 1 0 1 1 1 1 1 5F SP 0 0 1 0 0 0 0 0 20 ` 1 1 1 0 0 0 0 0 60 ! 1 0 1 0 0 0 0 1 21 a 0 1 1 0 0 0 0 1 61 ″ 1 0 1 0 0 0 1 0 22 b 0 1 1 0 0 0 1 0 62 # 0 0 1 0 0 0 1 1 23 c 1 1 1 0 0 0 1 1 63 $ 1 0 1 0 0 1 0 0 24 d 0 1 1 0 0 1 0 0 64 % 0 0 1 0 0 1 0 1 25 e 1 1 1 0 0 1 0 1 65 0 0 1 0 0 1 1 0 26 f 1 1 1 0 0 1 1 0 66 ′ 1 0 1 0 0 1 1 1 27 g 0 1 1 0 0 1 1 1 67 ( 1 0 1 0 1 0 0 0 28 h 0 1 1 0 1 0 0 0 68 ) 0 0 1 0 1 0 0 1 29 i 1 1 1 0 1 0 0 1 69 * 0 0 1 0 1 0 1 0 2A j 1 1 1 0 1 0 1 0 6A 1 0 1 0 1 0 1 1 2B k 0 1 1 0 1 0 1 1 6B , 0 0 1 0 1 1 0 0 2C l 1 1 1 0 1 1 0 0 6C - 1 0 1 0 1 1 0 1 2D m 0 1 1 0 1 1 0 1 6D . 1 0 1 0 1 1 1 0 2E n 0 1 1 0 1 1 1 0 6E / 0 0 1 0 1 1 1 1 2F o 1 1 1 0 1 1 1 1 6F 0 1 0 1 1 0 0 0 0 30 p 0 1 1 1 0 0 0 0 70 1 0 0 1 1 0 0 0 1 31 q 1 1 1 1 0 0 0 1 71 2 0 0 1 1 0 0 1 0 32 r 1 1 1 1 0 0 1 0 72 3 1 0 1 1 0 0 1 1 33 s 0 1 1 1 0 0 1 1 73 4 0 0 1 1 0 1 0 0 34 t 1 1 1 1 0 1 0 0 74 5 1 0 1 1 0 1 0 1 35 u 0 1 1 1 0 1 0 1 75 6 1 0 1 1 0 1 1 0 36 v 0 1 1 1 0 1 1 0 76 7 0 0 1 1 0 1 1 1 37 w 1 1 1 1 0 1 1 1 77 8 0 0 1 1 1 0 0 0 38 x 1 1 1 1 1 0 0 0 78 (Continued) 152
- 158. Fundamental Concepts of Data Communications Table 2 (Continued) Binary Code Binary Code Bit 7 6 5 4 3 2 1 0 Hex Bit 7 6 5 4 3 2 1 0 Hex 9 1 0 1 1 1 0 0 1 39 y 0 1 1 1 1 0 0 1 79 : 1 0 1 1 1 0 1 0 3A z 0 1 1 1 1 0 1 0 7A ; 0 0 1 1 1 0 1 1 3B { 1 1 1 1 1 0 1 1 7B 1 0 1 1 1 1 0 0 3C | 0 1 1 1 1 1 0 0 7C 0 0 1 1 1 1 0 1 3D } 1 1 1 1 1 1 0 1 7D 0 0 1 1 1 1 1 0 3E ⬃ 1 1 1 1 1 1 1 0 7E ? 1 0 1 1 1 1 1 1 3F DEL 0 1 1 1 1 1 1 1 7F NUL null VT vertical tab SYN synchronous SOH start of heading FF form feed ETB end of transmission block STX start of text CR carriage return CAN cancel ETX end of text SO shift-out SUB substitute EOT end of transmission SI shift-in ESC escape ENQ enquiry DLE data link escape FS field separator ACK acknowledge DC1 device control 1 GS group separator BEL bell DC2 device control 2 RS record separator BS back space DC3 device control 3 US unit separator HT horizontal tab DC4 device control 4 SP space NL new line NAK negative acknowledge DEL delete Table 3 EBCDIC Code Binary Code Binary Code Bit 0 1 2 3 4 5 6 7 Hex Bit 0 1 2 3 4 5 6 7 Hex NUL 0 0 0 0 0 0 0 0 00 1 0 0 0 0 0 0 0 80 SOH 0 0 0 0 0 0 0 1 01 a 1 0 0 0 0 0 0 1 81 STX 0 0 0 0 0 0 1 0 02 b 1 0 0 0 0 0 1 0 82 ETX 0 0 0 0 0 0 1 1 03 c 1 0 0 0 0 0 1 1 83 0 0 0 0 0 1 0 0 04 d 1 0 0 0 0 1 0 0 84 PT 0 0 0 0 0 1 0 1 05 e 1 0 0 0 0 1 0 1 85 0 0 0 0 0 1 1 0 06 f 1 0 0 0 0 1 1 0 86 0 0 0 0 0 1 1 1 07 g 1 0 0 0 0 1 1 1 87 0 0 0 0 1 0 0 0 08 h 1 0 0 0 1 0 0 0 88 0 0 0 0 1 0 0 1 09 i 1 0 0 0 1 0 0 1 89 0 0 0 0 1 0 1 0 0A 1 0 0 0 1 0 1 0 8A 0 0 0 0 1 0 1 1 0B 1 0 0 0 1 0 1 1 8B FF 0 0 0 0 1 1 0 0 0C 1 0 0 0 1 1 0 0 8C 0 0 0 0 1 1 0 1 0D 1 0 0 0 1 1 0 1 8D 0 0 0 0 1 1 1 0 0E 1 0 0 0 1 1 1 0 8E 0 0 0 0 1 1 1 1 0F 1 0 0 0 1 1 1 1 8F DLE 0 0 0 1 0 0 0 0 10 1 0 0 1 0 0 0 0 90 SBA 0 0 0 1 0 0 0 1 11 j 1 0 0 1 0 0 0 1 91 EUA 0 0 0 1 0 0 1 0 12 k 1 0 0 1 0 0 1 0 92 IC 0 0 0 1 0 0 1 1 13 l 1 0 0 1 0 0 1 1 93 0 0 0 1 0 1 0 0 14 m 1 0 0 1 0 1 0 0 94 NL 0 0 0 1 0 1 0 1 15 n 1 0 0 1 0 1 0 1 95 0 0 0 1 0 1 1 0 16 o 1 0 0 1 0 1 1 0 96 0 0 0 1 0 1 1 1 17 p 1 0 0 1 0 1 1 1 97 0 0 0 1 1 0 0 0 18 q 1 0 0 1 1 0 0 0 98 EM 0 0 0 1 1 0 0 1 19 r 1 0 0 1 1 0 0 1 99 0 0 0 1 1 0 1 0 1A 1 0 0 1 1 0 1 0 9A 0 0 0 1 1 0 1 1 1B 1 0 0 1 1 0 1 1 9B DUP 0 0 0 1 1 1 0 0 1C 1 0 0 1 1 1 0 0 9C SF 0 0 0 1 1 1 0 1 1D 1 0 0 1 1 1 0 1 9D FM 0 0 0 1 1 1 1 0 1E 1 0 0 1 1 1 1 0 9E (Continued) 153
- 159. Fundamental Concepts of Data Communications (Continued) Binary Code Binary Code Bit 0 1 2 3 4 5 6 7 Hex Bit 0 1 2 3 4 5 6 7 Hex ITB 0 0 0 1 1 1 1 1 1F 1 0 0 1 1 1 1 1 9F 0 0 1 0 0 0 0 0 20 1 0 1 0 0 0 0 0 A0 0 0 1 0 0 0 0 1 21 ⬃ 1 0 1 0 0 0 0 1 A1 0 0 1 0 0 0 1 0 22 s 1 0 1 0 0 0 1 0 A2 0 0 1 0 0 0 1 1 23 t 1 0 1 0 0 0 1 1 A3 0 0 1 0 0 1 0 0 24 u 1 0 1 0 0 1 0 0 A4 0 0 1 0 0 1 0 1 25 v 1 0 1 0 0 1 0 1 A5 ETB 0 0 1 0 0 1 1 0 26 w 1 0 1 0 0 1 1 0 A6 ESC 0 0 1 0 0 1 1 1 27 x 1 0 1 0 0 1 1 1 A7 0 0 1 0 1 0 0 0 28 y 1 0 1 0 1 0 0 0 A8 0 0 1 0 1 0 0 1 29 z 1 0 1 0 1 0 0 1 A9 0 0 1 0 1 0 1 0 2A 1 0 1 0 1 0 1 0 AA 0 0 1 0 1 0 1 1 2B 1 0 1 0 1 0 1 1 AB 0 0 1 0 1 1 0 0 2C 1 0 1 0 1 1 0 0 AC ENQ 0 0 1 0 1 1 0 1 2D 1 0 1 0 1 1 0 1 AD 0 0 1 0 1 1 1 0 2E 1 0 1 0 1 1 1 0 AE 0 0 1 0 1 1 1 1 2F 1 0 1 0 1 1 1 1 AF 0 0 1 1 0 0 0 0 30 1 0 1 1 0 0 0 0 B0 0 0 1 1 0 0 0 1 31 1 0 1 1 0 0 0 1 B1 SYN 0 0 1 1 0 0 1 0 32 1 0 1 1 0 0 1 0 B2 0 0 1 1 0 0 1 1 33 1 0 1 1 0 0 1 1 B3 0 0 1 1 0 1 0 0 34 1 0 1 1 0 1 0 0 B4 0 0 1 1 0 1 0 1 35 1 0 1 1 0 1 0 1 B5 0 0 1 1 0 1 1 0 36 1 0 1 1 0 1 1 0 B6 BOT 0 0 1 1 0 1 1 1 37 1 0 1 1 0 1 1 1 B7 0 0 1 1 1 0 0 0 38 1 0 1 1 1 0 0 0 B8 0 0 1 1 1 0 0 1 39 1 0 1 1 1 0 0 1 B9 0 0 1 1 1 0 1 0 3A 1 0 1 1 1 0 1 0 BA 0 0 1 1 1 0 1 1 3B 1 0 1 1 1 0 1 1 BB RA 0 0 1 1 1 1 0 0 3C 1 0 1 1 1 1 0 0 BC NAK 0 0 1 1 1 1 0 1 3D 1 0 1 1 1 1 0 1 BD 0 0 1 1 1 1 1 0 3E 1 0 1 1 1 1 1 0 BE SUB 0 0 1 1 1 1 1 1 3F 1 0 1 1 1 1 1 1 BF SP 0 1 0 0 0 0 0 0 40 { 1 1 0 0 0 0 0 0 C0 0 1 0 0 0 0 0 1 41 A 1 1 0 0 0 0 0 1 C1 0 1 0 0 0 0 1 0 42 B 1 1 0 0 0 0 1 0 C2 0 1 0 0 0 0 1 1 43 C 1 1 0 0 0 0 1 1 C3 0 1 0 0 0 1 0 0 44 D 1 1 0 0 0 1 0 0 C4 0 1 0 0 0 1 0 1 45 E 1 1 0 0 0 1 0 1 C5 0 1 0 0 0 1 1 0 46 F 1 1 0 0 0 1 1 0 C6 0 1 0 0 0 1 1 1 47 G 1 1 0 0 0 1 1 1 C7 0 1 0 0 1 0 0 0 48 H 1 1 0 0 1 0 0 0 C8 0 1 0 0 1 0 0 1 49 I 1 1 0 0 1 0 0 1 C9 ¢ 0 1 0 0 1 0 1 0 4A 1 1 0 0 1 0 1 0 CA - 0 1 0 0 1 0 1 1 4B 1 1 0 0 1 0 1 1 CB 0 1 0 0 1 1 0 0 4C 1 1 0 0 1 1 0 0 CC ( 0 1 0 0 1 1 0 1 4D 1 1 0 0 1 1 0 1 CD 0 1 0 0 1 1 1 0 4E 1 1 0 0 1 1 1 0 CE | 0 1 0 0 1 1 1 1 4F 1 1 0 0 1 1 1 1 CF 0 1 0 1 0 0 0 0 50 } 1 1 0 1 0 0 0 0 D0 0 1 0 1 0 0 0 1 51 J 1 1 0 1 0 0 0 1 D1 0 1 0 1 0 0 1 0 52 K 1 1 0 1 0 0 1 0 D2 0 1 0 1 0 0 1 1 53 L 1 1 0 1 0 0 1 1 D3 0 1 0 1 0 1 0 0 54 M 1 1 0 1 0 1 0 0 D4 0 1 0 1 0 1 0 1 55 N 1 1 0 1 0 1 0 1 D5 0 1 0 1 0 1 1 0 56 O 1 1 0 1 0 1 1 0 D6 (Continued) Table 3 154
- 160. Fundamental Concepts of Data Communications Table 3 (Continued) Binary Code Binary Code Bit 0 1 2 3 4 5 6 7 Hex Bit 0 1 2 3 4 5 6 7 Hex 0 1 0 1 0 1 1 1 57 P 1 1 0 1 0 1 1 1 D7 0 1 0 1 1 0 0 0 58 Q 1 1 0 1 1 0 0 0 D8 0 1 0 1 1 0 0 1 59 R 1 1 0 1 1 0 0 1 D9 ! 0 1 0 1 1 0 1 0 5A 1 1 0 1 1 0 1 0 DA $ 0 1 0 1 1 0 1 1 5B 1 1 0 1 1 0 1 1 DB * 0 1 0 1 1 1 0 0 5C 1 1 0 1 1 1 0 0 DC ) 0 1 0 1 1 1 0 1 5D 1 1 0 1 1 1 0 1 DD : 0 1 0 1 1 1 1 0 5E 1 1 0 1 1 1 1 0 DE ¬ 0 1 0 1 1 1 1 1 5F 1 1 0 1 1 1 1 1 DF 0 1 1 0 0 0 0 0 60 1 1 1 0 0 0 0 0 E0 / 0 1 1 0 0 0 0 1 61 1 1 1 0 0 0 0 1 E1 0 1 1 0 0 0 1 0 62 S 1 1 1 0 0 0 1 0 E2 0 1 1 0 0 0 1 1 63 T 1 1 1 0 0 0 1 1 E3 0 1 1 0 0 1 0 0 64 U 1 1 1 0 0 1 0 0 E4 0 1 1 0 0 1 0 1 65 V 1 1 1 0 0 1 0 1 E5 0 1 1 0 0 1 1 0 66 W 1 1 1 0 0 1 1 0 E6 0 1 1 0 0 1 1 1 67 X 1 1 1 0 0 1 1 1 E7 0 1 1 0 1 0 0 0 68 Y 1 1 1 0 1 0 0 0 E8 0 1 1 0 1 0 0 1 69 Z 1 1 1 0 1 0 0 1 E9 0 1 1 0 1 0 1 0 6A 1 1 1 0 1 0 1 0 EA 0 1 1 0 1 0 1 1 6B 1 1 1 0 1 0 1 1 EB % 0 1 1 0 1 1 0 0 6C 1 1 1 0 1 1 0 0 EC 0 1 1 0 1 1 0 1 6D 1 1 1 0 1 1 0 1 ED 0 1 1 0 1 1 1 0 6E 1 1 1 0 1 1 1 0 EE ? 0 1 1 0 1 1 1 1 6F 1 1 1 0 1 1 1 1 EF 0 1 1 1 0 0 0 0 70 0 1 1 1 1 0 0 0 0 F0 0 1 1 1 0 0 0 1 71 1 1 1 1 1 0 0 0 1 F1 0 1 1 1 0 0 1 0 72 2 1 1 1 1 0 0 1 0 F2 0 1 1 1 0 0 1 1 73 3 1 1 1 1 0 0 1 1 F3 0 1 1 1 0 1 0 0 74 4 1 1 1 1 0 1 0 0 F4 0 1 1 1 0 1 0 1 75 5 1 1 1 1 0 1 0 1 F5 0 1 1 1 0 1 1 0 76 6 1 1 1 1 0 1 1 0 F6 0 1 1 1 0 1 1 1 77 7 1 1 1 1 0 1 1 1 F7 0 1 1 1 1 0 0 0 78 8 1 1 1 1 1 0 0 0 F8 䊱 0 1 1 1 1 0 0 1 79 9 1 1 1 1 1 0 0 1 F9 : 0 1 1 1 1 0 1 0 7A 1 1 1 1 1 0 1 0 FA # 0 1 1 1 1 0 1 1 7B 1 1 1 1 1 0 1 1 FB @ 0 1 1 1 1 1 0 0 7C 1 1 1 1 1 1 0 0 FC 䊱 0 1 1 1 1 1 0 1 7D 1 1 1 1 1 1 0 1 FD 0 1 1 1 1 1 1 0 7E 1 1 1 1 1 1 1 0 FE ” 0 1 1 1 1 1 1 1 7F 1 1 1 1 1 1 1 1 FF DLE data-link escape ITB end of intermediate transmission block DUP duplicate NUL null EM end of medium PT program tab ENQ enquiry RA repeat to address EOT end of transmission SBA set buffer address ESC escape SF start field ETB end of transmission block SOH start of heading ETX end of text SP space EUA erase unprotected to address STX start of text FF form feed SUB substitute FM field mark SYN synchronous IC insert cursor NAK negative acknowledge 155
- 161. Fundamental Concepts of Data Communications FIGURE 1 Typical bar code 3 BAR CODES Bar codes are those omnipresent black-and-white striped stickers that seem to appear on virtually every consumer item in the United States and most of the rest of the world. Al- though bar codes were developed in the early 1970s, they were not used extensively un- til the mid-1980s. A bar code is a series of vertical black bars separated by vertical white bars (called spaces). The widths of the bars and spaces along with their reflective abili- ties represent binary 1s and 0s, and combinations of bits identify specific items. In addi- tion, bar codes may contain information regarding cost, inventory management and con- trol, security access, shipping and receiving, production counting, document and order processing, automatic billing, and many other applications. A typical bar code is shown in Figure 1. There are several standard bar code formats. The format selected depends on what types of data are being stored, how the data are being stored, system performance, and which format is most popular with business and industry. Bar codes are generally classified as being discrete, continuous, or two-dimensional (2D). Discrete code. A discrete bar code has spaces or gaps between characters. Therefore, each character within the bar code is independent of every other character. Code 39 is an example of a discrete bar code. Continuous code. A continuous bar code does not include spaces between characters. An example of a continuous bar code is the Universal Product Code (UPC). 2D code. A 2D bar code stores data in two dimensions in contrast with a conventional linear bar code, which stores data along only one axis. 2D bar codes have a larger storage capacity than one-dimensional bar codes (typically 1 kilobyte or more per data symbol). 3-1 Code 39 One of the most popular bar codes was developed in 1974 and called Code 39 (also called Code 3 of 9 and 3 of 9 Code). Code 39 uses an alphanumeric code similar to the ASCII code. Code 39 is shown in Table 4. Code 39 consists of 36 unique codes representing the 10 digits and 26 uppercase letters. There are seven additional codes used for special char- acters, and an exclusive start/stop character coded as an asterisk (*). Code 39 bar codes are ideally suited for making labels, such as name badges. Each Code 39 character contains nine vertical elements (five bars and four spaces). The logic condition (1 or 0) of each element is encoded in the width of the bar or space (i.e., width modulation). A wide element, whether it be a bar or a space, represents a logic 1, and a narrow element represents a logic 0. Three of the nine elements in each Code 39 character must be logic 1s, and the rest must be logic 0s. In addition, of the three logic 1s, two must be bars and one a space. Each character begins and ends with a black bar with alternating white bars in between. Since Code 39 is a discrete code, all characters are separated with an intercharacter gap, which is usually one character wide. The aster- isks at the beginning and end of the bar code are start and stop characters, respectively. 156
- 162. Fundamental Concepts of Data Communications Figure 2 shows the Code 39 representation of the start/stop code (*) followed by an in- tercharacter gap and then the Code 39 representation of the letter A. 3-2 Universal Product Code The grocery industry developed the Universal Product Code (UPC) sometime in the early 1970s to identify their products. The National Association of Food Chains officially adopted the UPC code in 1974. Today UPC codes are found on virtually every grocery item from a candy bar to a can of beans. Table 4 Code 39 Character Set Character Binary Code Bars Spaces Check Sum b8 b7 b6 b5 b4 b3 b2 b1 b0 b8b6b4b2b0 b7b5b3b1 Value 0 0 0 0 1 1 0 1 0 0 0 0 1 1 0 0 1 0 0 0 1 1 0 0 1 0 0 0 0 1 1 0 0 0 1 0 1 0 0 1 2 0 0 1 1 0 0 0 0 1 0 1 0 0 1 0 1 0 0 2 3 1 0 1 1 0 0 0 0 0 1 1 0 0 0 0 1 0 0 3 4 0 0 0 1 1 0 0 0 1 0 0 1 0 1 0 1 0 0 4 5 1 0 0 1 1 0 0 0 0 1 0 1 0 0 0 1 0 0 5 6 0 0 1 1 1 0 0 0 0 0 1 1 0 0 0 1 0 0 6 7 0 0 0 1 0 0 1 0 1 0 0 0 1 1 0 1 0 0 7 8 1 0 0 1 0 0 1 0 0 1 0 0 1 0 0 1 0 0 8 9 0 0 1 1 0 0 1 0 0 0 1 0 1 0 0 1 0 0 9 A 1 0 0 0 0 1 0 0 1 1 0 0 0 1 0 0 1 0 10 B 0 0 1 0 0 1 0 0 1 0 1 0 0 1 0 0 1 0 11 C 1 0 1 0 0 1 0 0 0 1 1 0 0 0 0 0 1 0 12 D 0 0 0 0 1 1 0 0 1 0 0 1 0 1 0 0 1 0 13 E 1 0 0 0 1 1 0 0 0 1 0 1 0 0 0 0 1 0 14 F 0 0 1 0 1 1 0 0 0 0 1 1 0 0 0 0 1 0 15 G 0 0 0 0 0 1 1 0 1 0 0 0 1 1 0 0 1 0 16 H 1 0 0 0 0 1 1 0 0 1 0 0 1 0 0 0 1 0 17 I 0 0 1 0 0 1 1 0 0 0 1 0 1 0 0 0 1 0 18 J 0 0 0 0 1 1 1 0 0 0 0 1 1 0 0 0 1 0 19 K 1 0 0 0 0 0 0 1 1 1 0 0 0 1 0 0 0 1 20 L 0 0 1 0 0 0 0 1 1 0 1 0 0 1 0 0 0 1 21 M 1 0 1 0 0 0 0 1 0 1 1 0 0 0 0 0 0 1 22 N 0 0 0 0 1 0 0 1 1 0 0 1 0 1 0 0 0 1 23 O 1 0 0 0 1 0 0 1 0 1 0 1 0 0 0 0 0 1 24 P 0 0 1 0 1 0 0 1 0 0 1 1 0 0 0 0 0 1 25 Q 0 0 0 0 0 0 1 1 1 0 0 0 1 1 0 0 0 1 26 R 1 0 0 0 0 0 1 1 0 1 0 0 1 0 0 0 0 1 27 S 0 0 1 0 0 0 1 1 0 0 1 0 1 0 0 0 0 1 28 T 0 0 0 0 1 0 1 1 0 0 0 1 1 0 0 0 0 1 29 U 1 1 0 0 0 0 0 0 1 1 0 0 0 1 1 0 0 0 30 V 0 1 1 0 0 0 0 0 1 0 1 0 0 1 1 0 0 0 31 W 1 1 1 0 0 0 0 0 0 1 1 0 0 0 1 0 0 0 32 X 0 1 0 0 1 0 0 0 1 0 0 1 0 1 1 0 0 0 33 Y 1 1 0 0 1 0 0 0 0 1 0 1 0 0 1 0 0 0 34 Z 0 1 1 0 1 0 0 0 0 0 1 1 0 0 1 0 0 0 35 0 1 0 0 0 0 1 0 1 0 0 0 1 1 1 0 0 0 36 . 1 1 0 0 0 0 1 0 0 1 0 0 1 0 1 0 0 0 37 space 0 1 1 0 0 0 1 0 0 0 1 0 1 0 1 0 0 0 38 * 0 1 0 0 1 0 1 0 0 0 0 1 1 0 1 0 0 0 — $ 0 1 0 1 0 1 0 0 0 0 0 0 0 0 1 1 1 0 39 / 0 1 0 1 0 0 0 1 0 0 0 0 0 0 1 1 0 1 40 0 1 0 0 0 1 0 1 0 0 0 0 0 0 1 0 1 1 41 % 0 0 0 1 0 1 0 1 0 0 0 0 0 0 0 1 1 1 42 157
- 163. X 3X Bar code Binary code Character asterisk (*) Start guard pattern A Next character Intercharacter gap Intercharacter gap X = width of narrow bar or space 3X = width of wide bar or space 0 1 0 0 1 0 1 0 0 1 0 0 0 0 1 0 1 0 FIGURE 2 Code 39 bar code Figures 3a, b, and c show the character set, label format, and sample bit patterns for the standard UPC code. Unlike Code 39, the UPC code is a continuous code since there are no in- tercharacter spaces. Each UPC label contains a 12-digit number. The two long bars shown in Figure 3b on the outermost left- and right-hand sides of the label are called the start guard pat- tern and the stop guard pattern, respectively.The start and stop guard patterns consist of a 101 (bar-space-bar) sequence, which is used to frame the 12-digit UPC number. The left and right halves of the label are separated by a center guard pattern, which consists of two long bars in the center of the label (they are called long bars because they are physically longer than the other bars on the label). The two long bars are separated with a space between them and have spaces on both sides of the bars. Therefore, the UPC center guard pattern is 01010 as shown inFigure3b.ThefirstsixdigitsoftheUPCcodeareencodedonthelefthalfofthelabel(called the left-hand characters), and the last six digits of the UPC code are encoded on the right half (called the right-hand characters). Note in Figure 3a that there are two binary codes for each character.When a character appears in one of the first six digits of the code, it uses a left-hand code, and when a character appears in one of the last six digits, it uses a right-hand code. Note that the right-hand code is simply the complement of the left-hand code. For example, if the second and ninth digits of a 12-digit code UPC are both 4s, the digit is encoded as a 0100011 in position 2 and as a 1011100 in position 9. The UPC code for the 12-digit code 012345 543210 is Fundamental Concepts of Data Communications 0001101 0011001 0010011 0111101 0100011 1011100 0110001 1001110 1000010 1101100 1100110 1110010 0 1 2 3 4 5 5 4 3 2 1 0 left-hand codes right-hand codes The first left-hand digit in the UPC code is called the UPC number system character, as it identifies how the UPC symbol is used. Table 5 lists the 10 UPC number system char- acters. For example, the UPC number system character 5 indicates that the item is intended to be used with a coupon. The other five left-hand characters are data characters. The first five right-hand characters are data characters, and the sixth right-hand character is a check character, which is used for error detection. The decimal value of the number system char- acter is always printed to the left of the UPC label, and on most UPC labels the decimal value of the check character is printed on the right side of the UPC label. With UPC codes, the width of the bars and spaces does not correspond to logic 1s and 0s. Instead, the digits 0 through 9 are encoded into a combination of two variable- 158
- 164. Fundamental Concepts of Data Communications Left-hand character Decimal digit Right-hand character 0001101 0011001 0010011 0111101 0100011 0110001 0101111 0111011 0110111 0001011 0 1 2 3 4 5 6 7 8 9 1110010 1100110 1101100 1000010 1011100 1001110 1010000 1000100 1001000 1110100 UPC Character Set (a) Number system character 101 101 01010 Center guard pattern Five left-hand data characters (35 bits) Five right-hand data characters (35 bits) 6 digits Start guard pattern Stop guard pattern Character check 6 digits (b) 0 1 0 0 0 1 1 Left-hand character 4 1 0 1 1 1 0 0 Right-hand character 4 (c) FIGURE 3 (a) UPC version A character set; (b) UPC label format; (c) left- and right-hand bit sequence for the digit 4 width bars and two variable-width spaces that occupy the equivalent of seven bit positions. Figure 3c shows the variable-width code for the UPC character 4 when used in one of the first six digit positions of the code (i.e., left-hand bit sequence) and when used in one of the last six digit positions of the code (i.e., right-hand bit sequence). A single bar (one bit po- sition) represents a logic 1, and a single space represents a logic 0. However, close exami- nation of the UPC character set in Table 5 will reveal that all UPC digits are comprised of bit patterns that yield two variable-width bars and two variable-width spaces, with the bar and space widths ranging from one to four bits. For the UPC character 4 shown in Figure 3c, the left-hand character is comprised of a one-bit space followed in order by a one-bit bar, a three-bit space, and a two-bit bar. The right-hand character is comprised of a one-bit bar followed in order by a one-bit space, a three-bit bar, and a two-bit space. 159
- 165. Table 5 UPC Number System Characters Character Intended Use 0 Regular UPC codes 1 Reserved for future use 2 Random-weight items that are symbol marked at the store 3 National Drug Code and National Health Related Items Code 4 Intended to be used without code format restrictions and with check digit protection for in-store marking of nonfood items 5 For use with coupons 6 Regular UPC codes 7 Regular UPC codes 8 Reserved for future use 9 Reserved for future use 0 0 0 1 1 0 1 Left-hand version of the character 0 Right-hand version of the character 0 1 1 1 0 0 1 0 FIGURE 4 UPC character 0 Example 1 Determine the UPC label structure for the digit 0. Solution From Figure 3a, the binary sequence for the digit 0 in the left-hand character field is 0001101, and the binary sequence for the digit 0 in the right-hand character field is 1110010. The left-hand sequence is comprised of three successive 0s, followed by two 1s, one 0, and one 1. The three successive 0s are equivalent to a space three bits long. The two 1s are equivalent to a bar two bits long. The single 0 and single 1 are equivalent to a space and a bar, each one bit long. The right-hand sequence is comprised of three 1s followed by two 0s, a 1, and a 0. The three 1s are equivalent to a bar three bits long. The two 0s are equivalent to a space two bits long. The sin- gle 1 and single 0 are equivalent to a bar and a space, each one bit long each. The UPC pattern for the digit 0 is shown in Figure 4. 4 ERROR CONTROL A data communications circuit can be as short as a few feet or as long as several thousand miles, and the transmission medium can be as simple as a pair of wires or as complex as a microwave, satellite, or optical fiber communications system. Therefore, it is inevitable that errors will occur, and it is necessary to develop and implement error-control procedures. Transmission errors are caused by electrical interference from natural sources, such as lightning, as well as from man-made sources, such as motors, generators, power lines, and fluorescent lights. Data communications errors can be generally classified as single bit, multiple bit, or burst. Single-bit errors are when only one bit within a given data string is in error. Single-bit errors affect only one character within a message. A multiple-bit error is when two or more nonconsecutive bits within a given data string are in error. Multiple-bit errors can affect one or more characters within a message.A burst error is when two or more consecutive bits within a given data string are in error. Burst errors can affect one or more characters within a message. Fundamental Concepts of Data Communications 160
- 166. Error performance is the rate in which errors occur, which can be described as either an expected or an empirical value. The theoretical (mathematical) expectation of the rate at which errors will occur is called probability of error (P[e]), whereas the actual historical record of a system’s error performance is called bit error rate (BER). For example, if a sys- tem has a P(e) of 105 , this means that mathematically the system can expect to experience one bit error for every 100,000 bits transported through the system (105 1/105 1/100,000). If a system has a BER of 105 , this means that in the past there was one bit er- ror for every 100,000 bits transported. Typically, a BER is measured and then compared with the probability of error to evaluate system performance. Error control can be divided into two general categories: error detection and error correction. 5 ERROR DETECTION Error detection is the process of monitoring data transmission and determining when errors have occurred. Error-detection techniques neither correct errors nor identify which bits are in error—they indicate only when an error has occurred. The purpose of error detection is not to prevent errors from occurring but to prevent undetected errors from occurring. The most common error-detection techniques are redundancy checking, which in- cludes vertical redundancy checking, checksum, longitudinal redundancy checking, and cyclic redundancy checking. 5-1 Redundancy Checking Duplicating each data unit for the purpose of detecting errors is a form of error detection called redundancy. Redundancy is an effective but rather costly means of detecting errors, especially with long messages. It is much more efficient to add bits to data units that check for transmission errors. Adding bits for the sole purpose of detecting errors is called redundancy checking. There are four basic types of redundancy checks: vertical redundancy checking, checksums, longitudinal redundancy checking, and cyclic redundancy checking. 5-1-1 Vertical redundancy checking. Vertical redundancy checking (VRC) is probably the simplest error-detection scheme and is generally referred to as character par- ity or simply parity. With character parity, each character has its own error-detection bit called the parity bit. Since the parity bit is not actually part of the character, it is considered a redundant bit. An n-character message would have n redundant parity bits. Therefore, the number of error-detection bits is directly proportional to the length of the message. With character parity, a single parity bit is added to each character to force the total number of logic 1s in the character, including the parity bit, to be either an odd number (odd parity) or an even number (even parity). For example, the ASCII code for the letter C is 43 hex, or P1000011 binary, where the P bit is the parity bit. There are three logic 1s in the code, not counting the parity bit. If odd parity is used, the P bit is made a logic 0, keeping the total number of logic 1s at three, which is an odd number. If even parity is used, the P bit is made a logic 1, making the total number of logic 1s four, which is an even number. The primary advantage of parity is its simplicity. The disadvantage is that when an even number of bits are received in error, the parity checker will not detect them because when the logic condition of an even number of bits is changed, the parity of the character remains the same. Consequently, over a long time, parity will theoretically detect only 50% of the transmission errors (this assumes an equal probability that an even or an odd number of bits could be in error). Example 2 Determine the odd and even parity bits for the ASCII character R. Solution The hex code for the ASCII character R is 52, which is P1010010 in binary, where P des- ignates the parity bit. Fundamental Concepts of Data Communications 161
- 167. For odd parity, the parity bit is a 0 because 52 hex contains three logic 1s, which is an odd num- ber. Therefore, the odd-parity bit sequence for the ASCII character R is 01010010. For even parity, the parity bit is 1, making the total number of logic 1s in the eight-bit sequence four, which is an even number. Therefore, the even-parity bit sequence for the ASCII character R is 11010010. Other forms of parity include marking parity (the parity bit is always a 1), no parity (the par- ity bit is not sent or checked), and ignored parity (the parity bit is always a 0 bit if it is ignored). Mark- ing parity is useful only when errors occur in a large number of bits. Ignored parity allows receivers that are incapable of checking parity to communicate with devices that use parity. 5-1-2 Checksum. Checksum is another relatively simple form of redundancy error checking where each character has a numerical value assigned to it. The characters within a message are combined together to produce an error-checking character (checksum), which can be as simple as the arithmetic sum of the numerical values of all the characters in the message. The checksum is appended to the end of the message. The receiver repli- cates the combining operation and determines its own checksum. The receiver’s checksum is compared to the checksum appended to the message, and if they are the same, it is as- sumed that no transmission errors have occurred. If the two checksums are different, a transmission error has definitely occurred. 5-1-3 Longitudinal redundancy checking. Longitudinal redundancy checking (LRC) is a redundancy error detection scheme that uses parity to determine if a transmis- sion error has occurred within a message and is therefore sometimes called message parity. With LRC, each bit position has a parity bit. In other words, b0 from each character in the message is XORed with b0 from all the other characters in the message. Similarly, b1, b2, and so on are XORed with their respective bits from all the characters in the message. Es- sentially, LRC is the result of XORing the “character codes” that make up the message, whereas VRC is the XORing of the bits within a single character. With LRC, even parity is generally used, whereas with VRC, odd parity is generally used. The LRC bits are computed in the transmitter while the data are being sent and then appended to the end of the message as a redundant character. In the receiver, the LRC is re- computed from the data, and the recomputed LRC is compared to the LRC appended to the message. If the two LRC characters are the same, most likely no transmission errors have occurred. If they are different, one or more transmission errors have occurred. Example 3 shows how are VRC and LRC are calculated and how they can be used to- gether. Example 3 Determine the VRCs and LRC for the following ASCII-encoded message: THE CAT. Use odd parity for the VRCs and even parity for the LRC. Solution Character T H E sp C A T LRC Hex 54 48 45 20 43 41 54 2F ASCII code b0 0 0 1 0 1 1 0 1 b1 0 0 0 0 1 0 0 1 b2 1 0 1 0 0 0 1 1 b3 0 1 0 0 0 0 0 1 b4 1 0 0 0 0 0 1 0 b5 0 0 0 1 0 0 0 1 b6 1 1 1 0 1 1 1 0 Parity bit b7 0 1 0 0 0 1 0 0 (VRC) Fundamental Concepts of Data Communications 162
- 168. The LRC is 00101111 binary (2F hex), which is the character “/” in ASCII. Therefore, after the LRC character is appended to the message, it would read “THE CAT/.” The group of characters that comprise a message (i.e., THE CAT) is often called a block or frame of data. Therefore, the bit sequence for the LRC is often called a block check sequence (BCS) or frame check sequence (FCS). With longitudinal redundancy checking, all messages (regardless of their length) have the same number of error-detection characters. This characteristic alone makes LRC a better choice for systems that typically send long messages. Historically, LRC detects between 95% and 98% of all transmission errors. LRC will not de- tect transmission errors when an even number of characters has an error in the same bit position. For example, if b4 in an even number of characters is in error, the LRC is still valid even though multiple transmission errors have occurred. 5-1-4 Cyclic redundancy checking. Probably the most reliable redundancy check- ing technique for error detection is a convolutional coding scheme called cyclic redundancy checking (CRC). With CRC, approximately 99.999% of all transmission errors are de- tected. In the United States, the most common CRC code is CRC-16. With CRC-16, 16 bits are used for the block check sequence. With CRC, the entire data stream is treated as a long continuous binary number. Because the BCS is separate from the message but transported within the same transmission, CRC is considered a systematic code. Cyclic block codes are often written as (n, k) cyclic codes where n bit length of transmission and k bit length of message. Therefore, the length of the BCC in bits is BCC n k A CRC-16 block check character is the remainder of a binary division process. A data message polynominal G(x) is divided by a unique generator polynominal function P(x), the quotient is discarded, and the remainder is truncated to 16 bits and appended to the message as a BCS. The generator polynominal must be a prime number (i.e., a number divisible by only itself and 1). CRC-16 detects all single-bit errors, all double- bit errors (provided the divisor contains at least three logic 1s), all odd number of bit errors (provided the division contains a factor 11), all error bursts of 16 bits or less, and 99.9% of error bursts greater than 16 bits long. For randomly distributed errors, it is es- timated that the likelihood of CRC-16 not detecting an error is 10 14 , which equates to one undetected error every two years of continuous data transmission at a rate of 1.544 Mbps. With CRC generation, the division is not accomplished with standard arithmetic di- vision. Instead, modulo-2 division is used, where the remainder is derived from an exclu- sive OR (XOR) operation. In the receiver, the data stream, including the CRC code, is di- vided by the same generating function P(x). If no transmission errors have occurred, the remainder will be zero. In the receiver, the message and CRC character pass through a block check register. After the entire message has passed through the register, its contents should be zero if the receive message contains no errors. Mathematically, CRC can be expressed as (1) where G(x) message polynominal P(x) generator polynominal Q(x) quotient R(x) remainder The generator polynomial for CRC-16 is P(x) x16 x15 x2 x0 G1x2 P1x2 Q1x2 R1x2 Fundamental Concepts of Data Communications 163
- 169. CRC-16 polynominal, G(x) = X16 + X15 + X2 + X0 X2 X15 X16 BCC output MSB LSB XOR XOR Data input 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 + + + XOR FIGURE 5 CRC-16 generating circuit The number of bits in the CRC code is equal to the highest exponent of the gener- ating polynomial. The exponents identify the bit positions in the generating polynomial that contain a logic 1. Therefore, for CRC-16, b16, b15, b2, and b0 are logic 1s, and all other bits are logic 0s. The number of bits in a CRC character is always twice the num- ber of bits in a data character (i.e., eight-bit characters use CRC-16, six-bit characters use CRC-12, and so on). Figure 5 shows the block diagram for a circuit that will generate a CRC-16 BCC. A CRC generating circuit requires one shift register for each bit in the BCC. Note that there are 16 shift registers in Figure 5. Also note that an XOR gate is placed at the output of the shift registers for each bit position of the generating polynomial that contains a logic 1, ex- cept for x0 . The BCC is the content of the 16 registers after the entire message has passed through the CRC generating circuit. Example 4 Determine the BCS for the following data and CRC generating polynomials: Data G(x) x7 x5 x4 x2 x1 x0 10110111 CRC P(x) x5 x4 x1 x0 110011 Solution First, G(x) is multiplied by the number of bits in the CRC code, which is 5: x5 (x7 x5 x4 x2 x1 x0 ) x12 x10 x9 x7 x6 x5 1011011100000 Then the result is divided by P(x): 1 1 0 1 0 1 1 1 1 1 0 0 1 1 | 1 0 1 1 0 1 1 1 0 0 0 0 0 1 1 0 0 1 1 1 1 1 1 0 1 1 1 0 0 1 1 1 1 1 0 1 0 1 1 0 0 1 1 1 0 0 1 0 0 1 1 0 0 1 1 1 0 1 1 1 0 1 1 0 0 1 1 1 1 1 0 1 0 1 1 0 0 1 1 0 1 0 0 1 CRC Fundamental Concepts of Data Communications 164
- 170. The CRC is appended to the data to give the following data stream: G(x) CRC 1 0 1 1 0 1 1 1 0 1 0 0 1 At the receiver, the data are again divided by P(x) : 1 1 0 1 0 1 1 1 1 1 0 0 1 1 | 1 0 1 1 0 1 1 1 0 1 0 0 1 1 1 0 0 1 1 1 1 1 1 0 1 1 1 0 0 1 1 1 1 1 0 1 0 1 1 0 0 1 1 1 0 0 1 1 0 1 1 0 0 1 1 1 0 1 0 1 0 1 1 0 0 1 1 1 1 0 0 1 1 1 1 0 0 1 1 0 0 0 0 0 0 Remainder 0, which means there were no transmis- sion errors 6 ERROR CORRECTION Although detecting errors is an important aspect of data communications, determining what to do with data that contain errors is another consideration. There are two basic types of er- ror messages: lost message and damaged message. A lost message is one that never arrives at the destination or one that arrives but is damaged to the extent that it is unrecognizable. A damaged message is one that is recognized at the destination but contains one or more transmission errors. Data communications network designers have developed two basic strategies for han- dling transmission errors: error-detecting codes and error-correcting codes. Error-de- tecting codes include enough redundant information with each transmitted message to en- able the receiver to determine when an error has occurred. Parity bits, block and frame check characters, and cyclic redundancy characters are examples of error-detecting codes. Error-correcting codes include sufficient extraneous information along with each message to enable the receiver to determine when an error has occurred and which bit is in error. Transmission errors can occur as single-bit errors or as bursts of errors, depending on the physical processes that caused them. Having errors occur in bursts is an advantage when data are transmitted in blocks or frames containing many bits. For example, if a typical frame size is 10,000 bits and the system has a probability of error of 104 (one bit error in every 10,000 bits transmitted), independent bit errors would most likely produce an error in every block. However, if errors occur in bursts of 1000, only one or two blocks out of every 1000 transmitted would contain errors. The disadvantage of bursts of errors is they are more difficult to detect and even more difficult to correct than isolated single-bit errors. In the modern world of data communications, there are two primary methods used for er- ror correction: retransmission and forward error correction. 6-1 Retransmission Retransmission,asthenameimplies,iswhenareceivestationrequeststhetransmitstationtore- send a message (or a portion of a message) when the message is received in error. Because the receive terminal automatically calls for a retransmission of the entire message, retransmission Fundamental Concepts of Data Communications 兵 兵 165
- 171. is often calledARQ, which is an old two-way radio term that means automatic repeat request or automaticretransmissionrequest.ARQisprobablythemostreliablemethodoferrorcorrection, although it is not necessarily the most efficient. Impairments on transmission media often occur in bursts. If short messages are used, the likelihood that impairments will occur during a trans- missionissmall.However,shortmessagesrequiremoreacknowledgmentsandlineturnarounds than do long messages.Acknowledgments are when the recipient of data sends a short message back to the sender acknowledging receipt of the last transmission.The acknowledgment can in- dicate a successful transmission (positive acknowledgment) or an unsuccessful transmission (negative acknowledgment). Line turnarounds are when a receive station becomes the transmit station, such as when acknowledgments are sent or when retransmissions are sent in response to a negative acknowledgment. Acknowledgments and line turnarounds for error control are forms of overhead (data other than user information that must be transmitted). With long mes- sages, less turnaround time is needed, although the likelihood that a transmission error will oc- cur is higher than for short messages. It can be shown statistically that messages between 256 and 512 characters long are the optimum size forARQ error correction. There are two basic types of ARQ: discrete and continuous. Discrete ARQ uses ac- knowledgments to indicate the successful or unsuccessful reception of data. There are two basic types of acknowledgments: positive and negative. The destination station responds with a positive acknowledgment when it receives an error-free message. The destination sta- tion responds with a negative acknowledgment when it receives a message containing er- rors to call for a retransmission. If the sending station does not receive an acknowledgment after a predetermined length of time (called a time-out), it retransmits the message. This is called retransmission after time-out. Another type of ARQ, called continuous ARQ, can be used when messages are di- vided into smaller blocks or frames that are sequentially numbered and transmitted in suc- cession, without waiting for acknowledgments between blocks. Continuous ARQ allows the destination station to asynchronously request the retransmission of a specific frame (or frames) of data and still be able to reconstruct the entire message once all frames have been successfully transported through the system. This technique is sometimes called selective repeat, as it can be used to call for a retransmission of an entire message or only a portion of a message. 6-2 Forward Error Correction Forward error correction (FEC) is the only error-correction scheme that actually detects and corrects transmission errors when they are received without requiring a retransmission. With FEC, redundant bits are added to the message before transmission. When an error is detected, the redundant bits are used to determine which bit is in error. Correcting the bit is a simple matter of complementing it. The number of redundant bits necessary to correct er- rors is much greater than the number of bits needed to simply detect errors. Therefore, FEC is generally limited to one-, two-, or three-bit errors. FEC is ideally suited for data communications systems when acknowledgments are impractical or impossible, such as when simplex transmissions are used to transmit mes- sages to many receivers or when the transmission, acknowledgment, and retransmission time is excessive, for example when communicating to far away places, such as deep-space vehicles. The purpose of FEC codes is to eliminate the time wasted for retransmissions. However, the addition of the FEC bits to each message wastes time itself. Obviously, a trade-off is made between ARQ and FEC, and system requirements determine which method is best suited to a particular application. Probably the most popular error-correction code is the Hamming code. 6-2-1 Hamming code. A mathematician named Richard W. Hamming, who was an early pioneer in the development of error-detection and -correction procedures, developed Fundamental Concepts of Data Communications 166
- 172. d1 d2 d3 d4 d5 d6 One data unit contains m + n bits m data bits n Hamming bits dm h1 h2 h3 hn FIGURE 6 Data unit comprised of m character bits and n Hamming bits the Hamming code while working at Bell Telephone Laboratories. The Hamming code is an error-correcting code used for correcting transmission errors in synchronous data streams. However, the Hamming code will correct only single-bit errors. It cannot correct multiple-bit errors or burst errors, and it cannot identify errors that occur in the Hamming bits themselves. The Hamming code, as with all FEC codes, requires the addition of over- head to the message, consequently increasing the length of a transmission. Hamming bits (sometimes called error bits) are inserted into a character at random lo- cations. The combination of the data bits and the Hamming bits is called the Hamming code. The only stipulation on the placement of the Hamming bits is that both the sender and the receiver must agree on where they are placed. To calculate the number of redundant Ham- ming bits necessary for a given character length, a relationship between the character bits and the Hamming bits must be established. As shown in Figure 6, a data unit contains m character bits and n Hamming bits. Therefore, the total number of bits in one data unit is m n. Since the Hamming bits must be able to identify which bit is in error, n Hamming bits must be able to indicate at least m n 1 different codes. Of the m n codes, one code in- dicates that no errors have occurred, and the remaining m n codes indicate the bit position where an error has occurred. Therefore, m n bit positions must be identified with n bits. Since n bits can produce 2n different codes, 2n must be equal to or greater than m n 1. Therefore, the number of Hamming bits is determined by the following expression: 2n ≥ m n 1 (2) where n number of Hamming bits m number of bits in each data character A seven-bit ASCII character requires four Hamming bits (24 7 4 1), which could be placed at the end of the character bits, at the beginning of the character bits, or in- terspersed throughout the character bits. Therefore, including the Hamming bits with ASCII-coded data requires transmitting 11 bits per ASCII character, which equates to a 57% increase in the message length. Example 5 For a 12-bit data string of 101100010010, determine the number of Hamming bits required, arbitrar- ily place the Hamming bits into the data string, determine the logic condition of each Hamming bit, assume an arbitrary single-bit transmission error, and prove that the Hamming code will successfully detect the error. Solution Substituting m 12 into Equation 2, the number of Hamming bits is for n 4 24 16 ≥ 12 4 1 17 Because 16 17, four Hamming bits are insufficient: for n 5 25 32 ≥ 12 5 1 18 Because 32 18, five Hamming bits are sufficient, and a total of 17 bits make up the data stream (12 data plus five Hamming). Fundamental Concepts of Data Communications 167
- 173. Arbitrarily placing five Hamming bits into bit positions 4, 8, 9, 13, and 17 yields bit position 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 H 1 0 1 H 1 0 0 H H 0 1 0 H 0 1 0 To determine the logic condition of the Hamming bits, express all bit positions that contain a logic 1 as a five-bit binary number and XOR them together: b17 1, b13 0, b9 1, b8 1, b4 0 The 17-bit Hamming code is H H H H H 1 1 0 1 0 1 0 0 1 1 0 1 0 0 0 1 0 Assume that during transmission, an error occurs in bit position 14. The received data stream is At the receiver, to determine the bit position in error, extract the Hamming bits and XOR them with the binary code for each data bit position that contains a logic 1: Therefore, bit position 14 contains an error. 7 CHARACTER SYNCHRONIZATION In essence, synchronize means to harmonize, coincide, or agree in time. Character syn- chronization involves identifying the beginning and end of a character within a message. When a continuous string of data is received, it is necessary to identify which bits belong to which characters and which bits are the MSBS and LSBS of the character. In essence, this is character synchronization: identifying the beginning and end of a character code. In data communications circuits, there are two formats commonly used to achieve character synchronization: asynchronous and synchronous. 7-1 Asynchronous Serial Data The term asynchronous literally means “without synchronism,” which in data communica- tions terminology means “without a specific time reference.” Asynchronous data transmis- Bit position Hamming bits 2 XOR 6 XOR 12 XOR 16 XOR Binary number 10110 00010 10100 00110 10010 01100 11110 10000 01110 14 1 1 0 0 0 1 0 0 1 1 0 1 0 0 0 1 0 error Bit position 2 6 XOR 12 XOR 14 XOR 16 XOR Binary number 00010 00110 00100 01100 01000 01110 00110 10000 10110 Hamming bits Fundamental Concepts of Data Communications 兵 168
- 174. sion is sometimes called start-stop transmission because each data character is framed be- tween start and stop bits. The start and stop bits identify the beginning and end of the char- acter, so the time gaps between characters do not present a problem. For asynchronously transmitted serial data, framing characters individually with start and stop bits is sometimes said to occur on a character-by-character basis. Figure 7 shows the format used to frame a character for asynchronous serial data transmission. The first bit transmitted is the start bit, which is always a logic 0. The char- acter bits are transmitted next, beginning with the LSB and ending with the MSB. The data character can contain between five and eight bits. The parity bit (if used) is transmitted di- rectly after the MSB of the character. The last bit transmitted is the stop bit, which is always a logic 1, and there can be either one, one and a half, or two stop bits. Therefore, a data char- acter may be comprised of between seven and 11 bits. A logic 0 is used for the start bit because an idle line condition (no data transmis- sion) on a data communications circuit is identified by the transmission of continuous logic 1s (called idle line 1s). Therefore, the start bit of a character is identified by a high- to-low transition in the received data, and the bit that immediately follows the start bit is the LSB of the character code. All stop bits are logic 1s, which guarantees a high-to-low transition at the beginning of each character.After the start bit is detected, the data and par- ity bits are clocked into the receiver. If data are transmitted in real time (i.e., as the opera- tor types data into the computer terminal), the number of idle line 1s between each char- acter will vary. During this dead time, the receive will simply wait for the occurrence of another start bit (i.e., high-to-low transition) before clocking in the next character. Obvi- ously, both slipping over and slipping under produce errors. However, the errors are some- what self-inflicted, as they occur in the receiver and are not a result of an impairment that occurred during transmission. With asynchronous data, it is not necessary that the transmit and receive clocks be continuously synchronized; however, their frequencies should be close, and they should be synchronized at the beginning of each character. When the transmit and receive clocks are substantially different, a condition called clock slippage may occur. If the transmit clock is substantially lower than the receive clock, underslipping occurs. If the transmit clock is substantially higher than the receive clock, a condition called overslipping occurs. With overslipping, the receive clock samples the receive data slower than the bit rate. Conse- quently, each successive sample occurs later in the bit time until finally a bit is completely skipped. Example 6 For the following sequence of bits, identify the ASCII-encoded character, the start and stop bits, and the parity bits (assume even parity and two stop bits): Fundamental Concepts of Data Communications (1) Stop bits (1, 1.5, or 2) Data bits (5 to 8) Start bit Data or Parity bit (odd/even) (1) 1 or 0 or b7 (MSB) b6 MSB b5 b4 b3 b2 b1 b0 LSB 0 Time FIGURE 7 Asynchronous data format 169
- 175. 1 1 1 1 1 1 0 1 0 0 0 0 0 1 0 1 1 1 1 0 0 0 1 0 0 0 1 1 1 1 1 1 0 1 0 0 0 0 0 1 0 1 1 1 1 0 0 0 1 0 0 0 one asynchronous character A 41 hex D 44 hex Stop bits Parity bit MSB MSB LSB LSB Parity bit Start bit Start bit Stop bits Idle line ones one asynchronous character time 7-2 Synchronous Serial Data Synchronous data generally involves transporting serial data at relatively high speeds in groups of characters called blocks or frames. Therefore, synchronous data are not sent in real time. Instead, a message is composed or formulated and then the entire message is transmitted as a single entity with no time lapses between characters. With synchronous data, rather than frame each character independently with start and stop bits, a unique se- quence of bits, sometimes called a synchronizing (SYN) character, is transmitted at the be- ginning of each message. For synchronously transmitted serial data, framing characters in blocks is sometimes said to occur on a block-by-block basis. For example, withASCII code, the SYN character is 16 hex. The receiver disregards incoming data until it receives one or more SYN characters. Once the synchronizing sequence is detected, the receiver clocks in the next eight bits and interprets them as the first character of the message. The receiver continues clocking in bits, interpreting them in groups of eight until it receives another unique character that signifies the end of the message. The end-of-message character varies with the type of protocol being used and what type of message it is associated with. With synchronous data, the transmit and receive clocks must be synchronized because character synchronization occurs only once at the beginning of a message. With synchronous data, each character has two or three bits added to each character (one start and either one, one and a half, or two stop bits). These bits are additional over- head and, thus, reduce the efficiency of the transmission (i.e., the ratio of information bits to total transmitted bits). Synchronous data generally has two SYN characters (16 bits of overhead) added to each message. Therefore, asynchronous data are more efficient for short messages, and synchronous data are more efficient for long messages. Example 7 For the following string of ASCII-encoded characters, identify each character (assume odd parity): Solution 0 1 0 0 1 1 1 1 0 1 0 1 0 1 0 0 0 0 0 1 0 1 1 0 1 1 0 1 0 0 1 1 1 1 0 1 0 1 0 1 0 0 0 0 0 1 0 1 1 0 1 1 Parity bit Parity bit MSB MSB MSB LSB LSB LSB Parity bit 4F hex O 54 hex T 16 hex SYN Character time Fundamental Concepts of Data Communications Solution 170
- 176. 8 DATA COMMUNICATIONS HARDWARE Digital information sources, such as personal computers, communicate with each other us- ing the POTS (plain old telephone system) telephone network in a manner very similar to the way analog information sources, such as human conversations, communicate with each other using the POTS telephone network. With both digital and analog information sources, special devices are necessary to interface the sources to the telephone network. Figure 8 shows a comparison between human speech (analog) communications and computer data (digital) communications using the POTS telephone network. Figure 8a shows how two humans communicate over the telephone network using standard analog telephone sets. The telephone sets interface human speech signals to the telephone network and vice versa. At the transmit end, the telephone set converts acoustical energy (information) to electrical energy and, at the receive end, the telephone set converts electrical energy back to acoustical energy. Figure 8b shows how digital data are transported over the telephone net- work. At the transmitting end, a telco interface converts digital data from the transceiver to analog electrical energy, which is transported through the telephone network. At the re- ceiving end, a telco interface converts the analog electrical energy received from the tele- phone network back to digital data. In simplified terms, a data communications system is comprised of three basic ele- ments: a transmitter (source), a transmission path (data channel), and a receiver (destina- tion). For two-way communications, the transmission path would be bidirectional and the source and destination interchangeable. Therefore, it is usually more appropriate to de- scribe a data communications system as connecting two endpoints (sometimes called nodes) through a common communications channel. The two endpoints may not possess the same computing capabilities; however, they must be configured with the same basic components. Both endpoints must be equipped with special devices that perform unique functions, make the physical connection to the data channel, and process the data before they are transmitted and after they have been received. Although the special devices are Fundamental Concepts of Data Communications Telephone network Acoustical energy Acoustical energy Human Telephone set Human Telephone set Electrical energy Electrical energy (a) Transceiver 1 Digital data Transceiver Digital data Telco interface Analog electrical energy Analog electrical energy Telco interface Telephone network (b) FIGURE 8 Telephone communications network: (a) human communications; (b) digital data communications 171
- 177. sometimes implemented as a single unit, it is generally easier to describe them as separate entities. In essence, all endpoints must have three fundamental components: data terminal equipment, data communications equipment, and a serial interface. 8-1 Data Terminal Equipment Data terminal equipment (DTE) can be virtually any binary digital device that generates, transmits, receives, or interprets data messages. In essence, a DTE is where information originates or terminates. DTEs are the data communications equivalent to the person in a telephone conversation. DTEs contain the hardware and software necessary to establish and control communications between endpoints in a data communications system; however, DTEs seldom communicate directly with other DTEs. Examples of DTEs include video display terminals, printers, and personal computers. Over the past 50 years, data terminal equipment has evolved from simple on-line printers to sophisticated high-level computers. Data terminal equipment includes the con- cept of terminals, clients, hosts, and servers. Terminals are devices used to input, output, and display information, such as keyboards, printers, and monitors. A client is basically a modern-day terminal with enhanced computing capabilities. Hosts are high-powered, high- capacity mainframe computers that support terminals. Servers function as modern-day hosts except with lower storage capacity and less computing capability. Servers and hosts maintain local databases and programs and distribute information to clients and terminals. 8-2 Data Communications Equipment Data communications equipment (DCE) is a general term used to describe equipment that in- terfaces data terminal equipment to a transmission channel, such as a digital T1 carrier or an analog telephone circuit. The output of a DTE can be digital or analog, depending on the ap- plication. In essence, a DCE is a signal conversion device, as it converts signals from a DTE to a form more suitable to be transported over a transmission channel. A DCE also converts those signals back to their original form at the receive end of a circuit. DCEs are transparent devices responsible for transporting bits (1s and 0s) between DTEs through a data communi- cations channel. The DCEs neither know nor do they care about the content of the data. There are several types of DCEs, depending on the type of transmission channel used. Common DCEs are channel service units (CSUs), digital service units (DSUs), and data modems. CSUs and DSUs are used to interface DTEs to digital transmission channels. Data modems are used to interface DTEs to analog telephone networks. Because data commu- nications channels are terminated at each end in a DCE, DCEs are sometimes called data circuit-terminating equipment (DCTE). Data modems are described in subsequent sections of this chapter. 9 DATA COMMUNICATIONS CIRCUITS A data modem is a DCE used to interface a DTE to an analog telephone circuit commonly called a POTS. Figure 9a shows a simplified diagram for a two-point data communications circuit using a POTS link to interconnect the two endpoints (endpoint A and endpoint B). As shown in the figure, a two-point data communications circuit is comprised of the seven basic components: 1. DTE at endpoint A 2. DCE at endpoint A 3. DTE/DCE interface at endpoint A 4. Transmission path between endpoint A and endpoint B 5. DCE at endpoint B 6. DTE at endpoint B 7. DTE/DCE interface at endpoint B Fundamental Concepts of Data Communications 172
- 178. The DTEs can be terminal devices, personal computers, mainframe computers, front- end processors, printers, or virtually any other piece of digital equipment. If a digital com- munications channel were used, the DCE would be a CSU or a DSU. However, because the communications channel is a POTS link, the DCE is a data modem. Figure 9b shows the same equivalent circuit as is shown in Figure 9a, except the DTE and DCE have been replaced with the actual devices they represent—the DTE is a personal computer, and the DCE is a modem. In most modern-day personal computers for home use, the modem is simply a card installed inside the computer. Figure 10 shows the block diagram for a centralized multipoint data communications circuit using several POTS data communications links to interconnect three endpoints. The circuit is arranged in a bus topology with central control provided by a mainframe computer (host) at endpoint A. The host station is sometimes called the primary station. Endpoints B and C are called secondary stations. The primary station is responsible for establishing and maintaining the data link and for ensuring an orderly flow of data between itself and each of the secondary stations. Data flow is controlled by an applications program stored in the mainframe computer at the primary station. At the primary station, there is a mainframe computer, a front-end processor (DTE), and a data modem (DCE).At each secondary station, there is a modem (DCE), a line control unit (DTE), and a cluster of terminal devices (personal computers, printers, and so on). The line control unit at the secondary stations is referred to as a cluster controller, as it controls data flow between several terminal devices and the data communications channel. Line con- trol units at secondary stations are sometimes called station controllers (STACOs), as they control data flow to and from all the data communications equipment located at that station. For simplicity, Figure 10 only shows one data circuit served by the mainframe com- puter at the primary station. However, there can be dozens of different circuits served by one mainframe computer. Therefore, the primary station line control unit (i.e., the front-end processor) must have enhanced capabilities for storing, processing, and retransmitting data it receives from all secondary stations on all the circuits it serves. The primary station stores software for database management of all the circuits it serves. Obviously, the duties Fundamental Concepts of Data Communications POTS Telephone network Transmission path DTE/DCE Interface DTE Endpoint A Endpoint B DTE DCE DCE DTE/DCE Interface (a) POTS Telephone network Transmission path DTE/DCE Interface Modem Modem PC Endpoint B PC Endpoint A DTE/DCE Interface (b) FIGURE 9 Two point data communications circuit: (a) DTE/DCE representation; (b) device representation 173
- 179. Modem (DCE) Front-end processor (DTE) Endpoint A Primary Station Endpoint B Secondary Mainframe computer (host) Parallel interface Modem (DCE) Transmission medium (POTS links) Line control unit (DTE) Printer Serial interface Serial interface PC1 Terminal devices PC2 Endpoint C Secondary Modem (DCE) Line control unit (DTE) Printer Serial interface PC1 Terminal devices PC2 FIGURE 10 Multipoint data communications circuit using POTS links performed by the front-end processor at the primary station are much more involved than the duties performed by the line control units at the secondary stations. The FEP directs data traffic to and from many different circuits, which could all have different parameters (i.e., different bit rates, character codes, data formats, protocols, and so on). The LCU at the sec- ondary stations directs data traffic between one data communications link and a relative few terminal devices, which all transmit and receive data at the same speed and use the same data-link protocol, character code, data format, and so on. 10 LINE CONTROL UNIT As previously stated, a line control unit (LCU) is a DTE, and DTEs have several important functions. At the primary station, the LCU is often called a FEP because it processes infor- mation and serves as an interface between the host computer and all the data communica- tions circuits it serves. Each circuit served is connected to a different port on the FEP. The FEP directs the flow of input and output data between data communications circuits and their respective application programs. The data interface between the mainframe computer and the FEP transfers data in parallel at relatively high bit rates. However, data transfers be- tween the modem and the FEP are accomplished in serial and at a much lower bit rate. The FEP at the primary station and the LCU at the secondary stations perform parallel-to-serial Fundamental Concepts of Data Communications 174
- 180. and serial-to-parallel conversions. They also house the circuitry that performs error detec- tion and correction. In addition, data-link control characters are inserted and deleted in the FEP and LCUs. Within the FEP and LCUs, a single special-purpose integrated circuit performs many of the fundamental data communications functions. This integrated circuit is called a universal asynchronous receiver/transmitter (UART) if it is designed for asynchronous data transmission, a universal synchronous receiver/transmitter (USRT) if it is designed for syn- chronous data transmission, and a universal synchronous/asynchronous receiver/transmitter (USART) if it is designed for either asynchronous or synchronous data transmission. All three types of circuits specify general-purpose integrated-circuit chips located in an LCU or FEP that allow DTEs to interface with DCEs. In modern-day integrated circuits, UARTs and USRTs are often combined into a single USART chip that is probably more popular to- day simply because it can be adapted to either asynchronous or synchronous data trans- mission. USARTs are available in 24- to 64-pin dual in-line packages (DIPs). UARTS, USRTS, and USARTS are devices that operate external to the central processor unit (CPU) in a DTE that allow the DTE to communicate serially with other data communications equipment, such as DCEs. They are also essential data communications components in terminals, workstations, PCs, and many other types of serial data commu- nications devices. In most modern computers, USARTs are normally included on the moth- erboard and connected directly to the serial port. UARTs, USRTs, and USARTs designed to interface to specific microprocessors often have unique manufacturer-specific names. For example, Motorola manufactures a special purpose UART chip it calls an asynchronous communications interface adapter (ACIA). 10-1 UART A UART is used for asynchronous transmission of serial data between a DTE and a DCE. Asynchronous data transmission means that an asynchronous data format is used, and there is no clocking information transferred between the DTE and the DCE. The primary func- tions performed by a UART are the following: 1. Parallel-to-serial data conversion in the transmitter and serial-to-parallel data con- version in the receiver 2. Error detection by inserting parity bits in the transmitter and checking parity bits in the receiver 3. Insert start and stop bits in the transmitter and detect and remove start and stop bits in the receiver 4. Formatting data in the transmitter and receiver (i.e., combining items 1 through 3 in a meaningful sequence) 5. Provide transmit and receive status information to the CPU 6. Voltage level conversion between the DTE and the serial interface and vice versa 7. Provide a means of achieving bit and character synchronization Transmit and receive functions can be performed by a UART simultaneously because the transmitter and receiver have separate control signals and clock signals and share a bidi- rectional data bus, which allows them to operate virtually independently of one another. In addition, input and output data are double buffered, which allows for continuous data trans- mission and reception. Figure 11 shows a simplified block diagram of a line control unit showing the rela- tionship between the UART and the CPU that controls the operation of the UART. The CPU coordinates data transfer between the line-control unit (or FEP) and the modem. The CPU is responsible for programming the UART’s control register, reading the UART’s status reg- ister, transferring parallel data to and from the UART transmit and receive buffer registers, providing clocking information to the UART, and facilitating the transfer of serial data be- tween the UART and the modem. Fundamental Concepts of Data Communications 175
- 181. UART Internal Registers Transmit buffer register Receive buffer register Control register Status word register Central processing unit (CPU) Address bus Transmit clock pulse TCP TBMT Transmit buffer empty Line-Control Unit (LCU) or Front-End Processor (FEP) Address Decoder CRS TSO RSI TDS SWE RDAR RDE To DCE From DCE Parallel data bus FIGURE 11 Line control unit UART interface 1 stop 1 stop Parity bit d6 data d5 data d4 data d3 data d2 data d1 data d0 data 0 start TSO TCP TCP CRS TEOC Serial data out to modem Data, parity, and stop-bit steering logic Control register TDS Transmit buffer register Parallel input data (5-8 bits) from CPU (TD7 through TD0) Control word inputs (8-bits) Transmit buffer register SWE TBMT Transmit shift register Transmit shift register empty logic circuit Status word register Output circuit FIGURE 12 UART transmitter block diagram Fundamental Concepts of Data Communications 176
- 182. A UART can be divided into two functional sections: the transmitter and the receiver. Figure 12 shows a simplified block diagram of a UART transmitter. Before transferring data in either direction, an eight-bit control word must be programmed into the UART control reg- ister to specify the nature of the data. The control word specifies the number of data bits per character; whether a parity bit is included with each character and, if so, whether it is odd or even parity; the number of stop bits inserted at the end of each character; and the receive clock frequency relative to the transmit clock frequency. Essentially, the start bit is the only bit in the UART that is not optional or programmable, as there is always one start bit, and it is al- ways a logic 0. Table 6 shows the control-register coding format for a typical UART. As specified in Table 6, the parity bit is optional and, if used, can be either odd or even. To select parity, NPB is cleared (logic 0), and to exclude the parity bit, NBP is set (logic 1). Odd parity is selected by clearing POE (logic 0), and even parity is selected by setting POE (logic 1).The number of stop bits is established with the NSB1 and NSB2 bits and can be one, one and a half, or two. The character length is determined by NDB1 and NDB2 and can be five, six, seven, or eight bits long. The maximum character length is 11 bits (i.e., one start bit, eight data bits, and two stop bits or one start bit, seven data bits, one parity bit, and two stop bits). Using a 22-bit character format with ASCII encoding is sometimes called full ASCII. Figure 13 shows three of the character formats possible with a UART. Figure 13a shows an 11-bit data character comprised of one start bit, seven ASCII data bits, one odd-parity bit, and two stop bits (i.e., full ASCII). Figure 13b shows a nine-bit data character com- prised of one start bit, seven ARQ data bits, and one stop bit, and Figure 13c shows another nine-bit data character comprised of one start bit, five Baudot data bits, one odd parity bit, and two stop bits. A UART also contains a status word register, which is an n-bit data register that keeps track of the status of the UART’s transmit and receive buffer registers. Typical status Fundamental Concepts of Data Communications Table 6 UART Control Register Inputs D7 and D6 Number of stop bits NSB1 NSB2 No. of Bits 0 0 Invalid 0 1 1 1 0 1.5 1 1 2 D5 and D4 NPB (parity or no parity) 1 No parity bit (RPE disabled in receiver) 0 Insert parity bits in transmitter and check parity bits in receiver POE (parity odd or even) 1 Even parity 0 Odd parity D3 and D2 Character length NDB1 NDB2 Bits per Word 0 0 5 0 1 6 1 0 7 1 1 8 D1 and D0 Receive clock (baud rate factor) RC1 RC2 Clock Rate 0 0 Synchronous mode 0 1 1X 1 0 16X 1 1 32X 177
- 183. 1 1 1 1 0 1 0 1 1 0 0 Two stop bits Start bit Odd parity ASCII Uppercase letter V (56 hex) 11-bit asynchronous ASCII character code (a) 1 1 0 1 0 0 0 1 0 One stop bit Start bit ARQ Uppercase letter M (51 hex) 9-bit asynchronous ARQ character code (b) 1 1 1 0 0 1 0 1 0 Two stop bits Odd parity Start bit Baudot Uppercase letter H (05 hex) 9-bit asynchronous Baudot character code (c) FIGURE 13 Asynchronous charac- ters: (a) ASCII character; (b) ARQ character; (c) Baudot character conditions compiled by the status word register for the UART transmitter include the fol- lowing conditions: TBMT: transmit buffer empty. Transmit shift register has completed transmission of a data character RPE: receive parity error. Set when a received character has a parity error in it RFE: receive framing error. Set when a character is received without any or with an improper number of stop bits ROR: receiver overrun. Set when a character in the receive buffer register is written over by another receive character because the CPU failed to service an active con- dition on REA before the next character was received from the receive shift register RDA: receive data available. A data character has been received and loaded into the receive data register 10-1-1 UART transmitter. The operation of the typical UART transmitter shown in Figure 12a is quite logical. However, before the UART can send or receive data, the UART control register must be loaded with the desired mode instruction word. This is ac- complished by the CPU in the DTE, which applies the mode instruction word to the con- trol word bus and then activates the control-register strobe (CRS). Figure 14 shows the signaling sequence that occurs between the CPU and the UART transmitter. On receipt of an active status word enable signal, the UART sends a transmit buffer empty (TBMT) signal from the status word register to the CPU to indicate that the transmit buffer register is empty and the UART is ready to receive more 1SWE2 Fundamental Concepts of Data Communications 178
- 184. Fundamental Concepts of Data Communications Central processing unit (CPU) UART Transmitter TEOC 1. SWE (status word enable) 2. TBMT (transmit buffer empty) 3. Parallel data (TD7-TD0) 4. TDS (transmit data strobe) 5. TSO (Transmit end of character) (Transmit serial data) FIGURE 14 UART transmitter signal sequence data. When the CPU senses an active condition of TBMT, it applies a parallel data char- acter to the transmit data lines (TD7 through TD0) and strobes them into the transmit buffer register with an active signal on the transmit data strobe signal. The con- tents of the transmit buffer register are transferred to the transmit shift register when the transmit end-of-character (TEOC) signal goes active (the TEOC signal is internal to the UART and simply tells the transmit buffer register when the transmit shift register is empty and available to receive data). The data pass through the steering logic circuit, where it picks up the appropriate start, stop, and parity bits. After data have been loaded into the transmit shift register, they are serially outputted on the transmit serial output (TSO) pin at a bit rate equal to the transmit clock (TCP) frequency. While the data in the transmit shift register are serially clocked out of the UART, the CPU applies the next char- acter to the input of the transmit buffer register. The process repeats until the CPU has transferred all its data. 10-1-2 UART receiver. A simplified block diagram for a UART receiver is shown in Figure 15. The number of stop bits and data bits and the parity bit parameters specified for the UART receiver must be the same as those of the UART transmitter. The UART re- ceiver ignores the reception of idle line 1s. When a valid start bit is detected by the start bit verification circuit, the data character is clocked into the receive shift register. If parity is used, the parity bit is checked in the parity checker circuit. After one complete data char- acter is loaded into the shift register, the character is transferred in parallel into the receive buffer register, and the receive data available (RDA) flag is set in the status word register. The CPU reads the status register by activating the status word enable signal and, if RDA is active, the CPU reads the character from the receive buffer register by placing an active signal on the receive data enable (RDE) pin. After reading the data, the CPU places an active signal on the receive data available reset pin, which resets the RDA pin. Meanwhile, the next character is received and clocked into the receive shift register, and the process repeats until all the data have been received. Figure 16 shows the receive signaling sequence that occurs between the CPU and the UART. 10-1-3 Start-bit verification circuit. With asynchronous data transmission, pre- cise timing is less important than following an agreed-on format or pattern for the data. Each transmitted data character must be preceded by a start bit and end with one or more 1RDAR2 1SWE2 1TDS2 179
- 185. Start bit verification circuit RSI input data from modem RCP (receive clock) Parity checker circuit Control register Receive shift register Receive buffer register RD7 RD6 RD5 RD4 RD3 RD2 RD1 RD0 Status word register RPE RFE RDA RDA RDE Parallel output data ROR SWE RDAR FIGURE 15 UART receiver block diagram Central processing unit (CPU) UART Receiver 3. SWE (status word enable) 5. SWE (status word enable) 7. RDE (receive data enable) 9. RDAR (receive data available reset) 1. Valid start bit detected 2. Receive data character loaded serially into receive shift register 4. RDA (receive data available) 6. Status word transfered to CPU 8. Parallel data (RD7-RD0) RSI serial data received from modem (RPE, RFE, and ROR) FIGURE 16 UART receive signal sequence Fundamental Concepts of Data Communications 180
- 186. stop bits. Because data received by a UART have been transmitted from a distant UART whose clock is asynchronous to the receive UART, bit synchronization is achieved by es- tablishing a timing reference at the center of each start bit. Therefore, it is imperative that a UART detect the occurrence of a valid start bit early in the bit cell and establish a timing reference before it begins to accept data. The primary function of the start bit verification circuit is to detect valid start bits, which indicate the beginning of a data character. Figure 17a shows an example of how a noise hit can be misinterpreted as a start bit. The input data consist of a continuous string Fundamental Concepts of Data Communications Idle line 1s 1X RCP Samples data high – idle line 1 Samples data detects low invalid start bit (a) Samples data interprets b0 as a logic 1 Samples data interprets b1 as a logic 1 Noise impulse Idle line 1s misinterpreted as data Idle line 1s 16X RCP Noise impulse wait 7 clock pulses before sampling again 1 2 3 4 Sample after 7 clock cycles high – invalid start bit (b) 5 6 7 Idle line 1s (ignored) Idle line 1s Start bit b0 b1 bit b1 = 0 bit b0 = 1 Sample again still low valid start bit Detects low 16X RCP Wait 16 clock pulses sample data Wait seven clock pulses Wait 16 clock pulses sample data FIGURE 17 Start bit verification: (a) 1X RCP; (b) 16X RCP; (c) valid start bit (c) 181
- 187. Idle line 1s Center of start bit Center of bit b0 Center of bit b1 bit b1 = 0 bit b0 = 1 Sample again still low valid start bit Sampling error Sampling error Sampling error Detects low 16X RCP Detection error Wait 16 clock pulses sample data Wait seven clock pulses Wait 16 clock pulses sample data FIGURE 18 16X receive clock rate of idle line 1s, which are typically transmitted when there is no information. Idle line 1s are interpreted by a receiver as continuous stop bits (i.e., no data). If a noise impulse occurs that causes the receive data to go low at the same time the receiver clock is active, the receiver will interpret the noise impulse as a start bit. If this happens, the receiver will misinterpret the logic condition present during the next clock as the first data bit (b0) and the follow- ing clock cycles as the remaining data bits (b1, b2, and so on). The likelihood of misinter- preting noise hits as start bits can be reduced substantially by clocking the UART receiver at a rate higher than the incoming data. Figure 17b shows the same situation as shown in Figure 17a, except the receive clock pulse (RCP) is 16 times (16) higher than the receive serial data input (RSI). Once a low is detected, the UART waits seven clock cycles before resampling the input data. Waiting seven clock cycles places the next sample very near the center of the start bit. If the next sample detects a low, it assumes that a valid start bit has been detected. If the data have reverted to the high condition, it is assumed that the high- to-low transition was simply a noise pulse and, therefore, is ignored. Once a valid start bit has been detected and verified (Figure 17c), the start bit verification circuit samples the in- coming data once every 16 clock cycles, which essentially makes the sample rate equal to the receive data rate (i.e., 16 RCP/16 RCP). The UART continues sampling the data once every 16 clock cycles until the stop bits are detected, at which time the start bit ver- ification circuit begins searching for another valid start bit. UARTs are generally pro- grammed for receive clock rates of 16, 32, or 64 times the receive data rate (i.e., 16, 32, and 64). Another advantage of clocking a UART receiver at a rate higher than the actual re- ceive data is to ensure that a high-to-low transition (valid start bit) is detected as soon as possible. This ensures that once the start bit is detected, subsequent samples will occur very near the center of each data bit. The difference in time between when a sample is taken (i.e., when a data bit is clocked into the receive shift register) and the actual center of a data bit is called the sampling error. Figure 18 shows a receive data stream sampled at a rate 16 times higher (16 RCP) than the actual data rate (RCP). As the figure shows, the start bit is not immediately detected. The difference in time between the beginning of a start bit and when it is detected is called the detection error. The maximum detection er- ror is equal to the time of one receive clock cycle (tcl 1/Rcl). If the receive clock rate equaled the receive data rate, the maximum detection error would approach the time of one bit, which would mean that a start bit would not be detected until the very end of the bit time. Obviously, the higher the receive clock rate, the earlier a start bit would be detected. Fundamental Concepts of Data Communications 182
- 188. Idle line 1s 8X 1 2 3 b0 Detects low Sampling error Center of start bit (Wait 3 clocks before sampling again) (a) Idle line 1s 16X 1 2 3 4 5 6 7 b0 Detects low Sampling error Center of start bit (b) (Wait 7 clocks before sampling again) FIGURE 19 Sampling error: (a) 8X RCP; (b) 16X RCP Because of the detection error, successive samples occur slightly off from the center of the data bit. This would not present a problem with synchronous clocks, as the sampling error would remain constant from one sample to the next. However, with asynchronous clocks, the magnitude of the sampling error for each successive sample would increase (the clock would slip over or slip under the data), eventually causing a data bit to be either sam- pled twice or not sampled at all, depending on whether the receive clock is higher or lower than the transmit clock. Figure 19 illustrates how sampling at a higher rate reduces the sampling error. Figures 19a and b show data sampled at a rate eight times the data rate (8) and 16 times the data rate (16), respectively. It can be seen that increasing the sample rate moves the sample time closer to the center of the data bit, thus decreasing the sampling error. Placing stop bits at the end of each data character also helps reduce the clock slip- page (sometimes called clock skew) problem inherent when using asynchronous trans- mit and receive clocks. Start and stop bits force a high-to-low transition at the beginning of each character, which essentially allows the receiver to resynchronize to the start bit at the beginning of each data character. It should probably be mentioned that with UARTs the data rates do not have to be the same in each direction of propagation (e.g., you could transmit data at 1200 bps and receive at 600 bps). However, the rate at which data leave a transmitter must be the same as the rate at which data enter the receiver at the other end of the circuit. If you transmit at 1200 bps, it must be received at the other end at 1200 bps. 10-2 Universal Synchronous Receiver/Transmitter A universal synchronous receiver/transmitter (USRT) is used for synchronous transmis- sion of data between a DTE and a DCE. Synchronous data transmission means that a syn- chronous data format is used, and clocking information is generally transferred between the DTE and the DCE. A USRT performs the same basic functions as a UART, except for Fundamental Concepts of Data Communications 183
- 189. synchronous data (i.e., the start and stop bits are omitted and replaced by unique synchro- nizing characters). The primary functions performed by a USRT are the following: 1. Serial-to-parallel and parallel-to-serial data conversions 2. Error detection by inserting parity bits in the transmitter and checking parity bits in the receiver. 3. Insert and detect unique data synchronization (SYN) characters 4. Formatting data in the transmitter and receiver (i.e., combining items 1 through 3 in a meaningful sequence) 5. Provide transmit and receive status information to the CPU 6. Voltage-level conversion between the DTE and the serial interface and vice versa 7. Provide a means of achieving bit and character synchronization 11 SERIAL INTERFACES To ensure an orderly flow of data between a DTE and a DCE, a standard serial interface is used to interconnect them. The serial interface coordinates the flow of data, control signals, and timing information between the DTE and the DCE. Before serial interfaces were standardized, every company that manufactured data communications equipment used a different interface configuration. More specifically, the cable arrangement between the DTE and the DCE, the type and size of the connectors, and the voltage levels varied considerably from vender to vender. To interconnect equipment manufactured by different companies, special level converters, cables, and connectors had to be designed, constructed, and implemented for each application. A serial interface stan- dard should provide the following: 1. A specific range of voltages for transmit and receive signal levels 2. Limitations for the electrical parameters of the transmission line, including source and load impedance, cable capacitance, and other electrical characteristics out- lined later in this chapter 3. Standard cable and cable connectors 4. Functional description of each signal on the interface In 1962, the Electronics Industries Association (EIA), in an effort to standardize inter- face equipment between data terminal equipment and data communications equipment, agreed on a set of standards called the RS-232 specifications (RS meaning “recommended standard”). The official name of the RS-232 interface is Interface Between Data Terminal Equipment and Data Communications Equipment Employing Serial Binary Data Inter- change. In 1969, the third revision, RS-232C, was published and remained the industrial stan- dard until 1987, when the RS-232D was introduced, which was followed by the RS-232E in theearly1990s.TheRS-232DstandardissometimesreferredtoastheEIA-232standard.Ver- sions D and E of the RS-232 standard changed some of the pin designations. For example, data set ready was changed to DCE ready, and data terminal ready was changed to DTE ready. The RS-232 specifications identify the mechanical, electrical, functional, and proce- dural descriptions for the interface between DTEs and DCEs. The RS-232 interface is sim- ilar to the combined ITU-T standards V.28 (electrical specifications) and V.24 (functional description) and is designed for serial transmission up to 20 kbps over a maximum distance of 50 feet (approximately 15 meters). 11-1 RS-232 Serial Interface Standard The mechanical specification for the RS-232 interface specifies a cable with two connec- tors. The standard RS-232 cable is a sheath containing 25 wires with a DB25P-compatible Fundamental Concepts of Data Communications 184
- 190. (a) (b) (c) (d) 25 13 24 23 22 21 20 19 18 17 16 15 14 12 25 13 11 10 9 8 7 6 5 4 3 2 1 (d) 1 2 3 4 6 7 8 9 5 FIGURE 20 RS-232 serial interface connector: (a) DB25P; (b) DB25S; (c) DB9P; (d) DB9S 1 (R) 2 (CD) 3 (DTR) 4 (SG) 5 (RD) 6 (TD) 7 (CTS) 8 (RTS) 1 2 3 4 5 6 7 8 FIGURE 21 EIA-561 modular connector male connector (plug) on one end and a DB25S-compatible female connector (receptacle) on the other end. The DB25P-compatible and DB25S-compatible connectors are shown in Figures 20a and b, respectively. The cable must have a plug on one end that connects to the DTE and a receptacle on the other end that connects to the DCE. There is also a spe- cial PC nine-pin version of the RS-232 interface cable with a DB9P-compatible male connector on one end and a DB9S-compatible connector at the other end. The DB9P- compatible and DB9S-compatible connectors are shown in Figures 20c and d, respec- tively (note that there is no correlation between the pin assignments for the two connec- tors). The nine-pin version of the RS-232 interface is designed for transporting asynchronous data between a DTE and a DCE or between two DTEs, whereas the 25-pin version is designed for transporting either synchronous or asynchronous data between a DTE and a DCE. Figure 21 shows the eight-pin EIA-561 modular connector, which is used for transporting asynchronous data between a DTE and a DCE when the DCE is con- nected directly to a standard two-wire telephone line attached to the public switched tele- phone network. The EIA-561 modular connector is designed exclusively for dial-up tele- phone connections. Although the RS-232 interface is simply a cable and two connectors, the standard also specifies limitations on the voltage levels that the DTE and DCE can output onto or receive from the cable. The DTE and DCE must provide circuits that convert their internal logic levelstoRS-232-compatiblevalues.Forexample,aDTEusingTTLlogicinterfacedtoaDCE using CMOS logic is not compatible. Voltage-leveling circuits convert the internal voltage levels from the DTE and DCE to RS-232 values. If both the DCE and the DTE output and acceptRS-232levels,theyareelectricallycompatibleregardlessofwhichlogicfamilytheyuse internally. A voltage leveler is called a driver if it outputs signals onto the cable and a Fundamental Concepts of Data Communications 185
- 191. Table 7 RS-232 Voltage Specifications Data Signals Control Signals Logic 1 Logic 0 Enable (On) Disable (Off) Driver (output) 5 V to 15 V 5 V to 15 V 5 V to 15 V 5 V to 15 V Terminator (input) 3 V to 25 V 3 V to 25 V 3 V to 25 V 3 V to 25 V terminator if it accepts signals from the cable. In essence, a driver is a transmitter, and a ter- minator is a receiver. Table 7 lists the voltage limits for RS-232-compatible drivers and ter- minators. Note that the data and control lines use non–return to zero, level (NRZ-L) bipolar encoding.However,thedatalinesusenegativelogic,whilethecontrollinesusepositivelogic. From examining Table 7, it can be seen that the voltage limits for a driver are more inclusive than the voltage limits for a terminator. The output voltage range for a driver is between 5 V and 15 V or between 5 V and 15 V, depending on the logic level. How- ever, the voltage range in which a terminator will accept is between 3 V and 25 V or be- tween 3 V and 25 V. Voltages between 3 V are undefined and may be interpreted by a terminator as a high or a low. The difference in the voltage levels between the driver out- put and the terminator input is called noise margin (NM). The noise margin reduces the sus- ceptibility to interface caused by noise transients induced into the cable. Figure 22a shows the relationship between the driver and terminator voltage ranges. As shown in Figure 22a, the noise margin for the minimum driver output voltage is 2 V (5 3), and the noise mar- gin for the maximum driver output voltage is 10 V (25 – 15). (The minimum noise margin of 2 V is called the implied noise margin.) Noise margins will vary, of course, depending on what specific voltages are used for highs and lows. When the noise margin of a circuit is a high value, it is said to have high noise immunity, and when the noise margin is a low value, it has low noise immunity. Typical RS-232 voltage levels are 10 V for a high and 10 V for a low, which produces a noise margin of 7 V in one direction and 15 V in the other direction. The noise margin is generally stated as the minimum value. This relation- ship is shown in Figure 22b. Figure 22c illustrates the immunity of the RS-232 interface to noise signals for logic levels of 10 V and 10 V. The RS-232 interface specifies single-end (unbalanced) operation with a common ground between the DTE and DCE. A common ground is reasonable when a short cable is used. However, with longer cables and when the DTE and DCE are powered from different electrical buses, this may not be true. Example 8 Determine the noise margins for an RS-232 interface with driver signal voltages of 6 V. Solution The noise margin is the difference between the driver signal voltage and the terminator receive voltage, or NM 6 3 3 V or NM 25 6 19 V The minimum noise margin is 3 V. 11-1-1 RS-232 electrical equivalent circuit. Figure 23 shows the equivalent elec- trical circuit for the RS-232 interface, including the driver and terminator. With these elec- trical specifications and for a bit rate of 20 kbps, the nominal maximum length of the RS- 232 interface cable is approximately 50 feet. 11-1-2 RS-232 functional description. The pins on the RS-232 interface cable are functionally categorized as either ground (signal and chassis), data (transmit and re- Fundamental Concepts of Data Communications 186
- 192. 10 V Noise margin 10 V Noise margin (a) 2 V Noise margin 2 V Noise margin 10 V Undefined Zone 6 V Undefined Zone RS-232 Terminator RS-232 Driver High High Low Low +25 V +3 V –3 V –25 V +15 V –15 V +5 V –5 V 0 V 15 V Noise margin 15 V Noise margin 7 V Noise margin 7 V Noise margin (b) 6 V Undefined Zone RS-232 Terminator RS-232 Driver High High Low Low +25 V +3 V –3 V –25 V +10 V –10 V 0 V FIGURE 22 RS-232 logic levels and noise margin: (a) driver and terminator voltage ranges; (b) noise margin with a 10 V high and 10 V low (Continued) ceive), control (handshaking and diagnostic), or timing (clocking signals). Although the RS-232 interface as a unit is bidirectional (signals propagate in both directions), each indi- vidual wire or pin is unidirectional. That is, signals on any given wire are propagated either from the DTE to the DCE or from the DCE to the DTE but never in both directions. Table 8 lists the 25 pins (wires) of the RS-232 interface and gives the direction of signal propa- gation (i.e., either from the DTE toward the DCE or from the DCE toward the DTE). The RS-232 specification designates the first letter of each pin with the letters A, B, C, D, or S. The letter categorizes the signal into one of five groups, each representing a different type of circuit. The five groups are as follows: A—ground B—data C—control D—timing (clocking) S—secondary channel Fundamental Concepts of Data Communications 187
- 193. RS-232 Connector RS-232 Connector RS-232 Interface cable Terminator Driver Rout CO Vi CL RL Vout Signal ground Vout — open-circuit voltage at the output of a driver (±5 V to ±15 V) Vi — terminated voltage at the input to a terminator (±3 V to ±25 V) CL — load capacitance associated with the terminator, including the cable (2500 pF maximum) CO — capacitance seen by the driver including the cable (2500 pF maximum) RL — terminator input resistance (3000 Ω to 7000 Ω) Rout — driver output resistance (300 Ω maximum) FIGURE 23 RS-232 Electrical specifications Fundamental Concepts of Data Communications Signal variations due to noise Signal variations due to noise Noise violation 6 V Undefined Zone RS-232 Terminator RS-232 Driver High High Low Low +25 V +3 V –3 V –25 V +10 V –10 V 0 V (c) FIGURE 22 (Continued) (c) noise violation 188
- 194. Because the letters are nondescriptive designations, it is more practical and useful to use acronyms to designate the pins that reflect the functions of the pins. Table 9 lists the EIA signal designations plus the nomenclature more commonly used by industry in the United States to designate the pins. Twenty of the 25 pins on the RS-232 interface are designated for specific purposes or functions. Pins 9, 10, 11, 18, and 25 are unassigned (unassigned does not necessarily imply unused). Pins 1 and 7 are grounds; pins 2, 3, 14, and 16 are data pins; pins 15, 17, and 24 are timing pins; and all the other pins are used for control or handshaking signals. Pins 1 through 8 are used with both asynchronous and synchronous modems. Pins 15, 17, and 24 are used only with synchronous modems. Pins 12, 13, 14, 16, and 19 are used only when the DCE is equipped with a secondary data channel. Pins 20 and 22 are used exclusively when interfacing a DTE to a modem that is connected to standard dial-up telephone circuits on the public switched telephone network. There are two full-duplex data channels available with the RS-232 interface; one channel is for primary data (actual information), and the second channel is for secondary data (diagnostic information and handshaking signals). The secondary chan- nel is sometimes used as a reverse or backward channel, allowing the receive DCE to communicate with the transmit DCE while data are being transmitted on the primary data channel. Fundamental Concepts of Data Communications Table 8 EIA RS-232 Pin Designations and Direction of Propagation Pin Number Pin Name Direction of Propagation 1 Protective ground (frame ground or chassis ground) None 2 Transmit data (send data) DTE to DCE 3 Receive data DCE to DTE 4 Request to send DTE to DCE 5 Clear to send DCE to DTE 6 Data set ready (modem ready) DCE to DTE 7 Signal ground (reference ground) None 8 Receive line signal detect (carrier detect or data carrier detect) DCE to DTE 9 Unassigned None 10 Unassigned None 11 Unassigned None 12 Secondary receive line signal detect (secondary carrier detect or secondary data carrier detect) DCE to DTE 13 Secondary clear to send DCE to DTE 14 Secondary transmit data (secondary send data) DTE to DCE 15 Transmit signal element timing—DCE (serial clock transmit—DCE) DCE to DTE 16 Secondary receive data DCE to DTE 17 Receive signal element timing (serial clock receive) DCE to DTE 18 Unassigned None 19 Secondary request to send DTE to DCE 20 Data terminal ready DTE to DCE 21 Signal quality detect DCE to DTE 22 Ring indicator DCE to DTE 23 Data signal rate selector DTE to DCE 24 Transmit signal element timing—DTE (serial clock transmit—DTE) DTE to DCE 25 Unassigned None 189
- 195. Table 9 EIA RS-232 Pin Designations and Designations Pin Pin EIA Common U.S. Number Name Nomenclature Acronyms 1 Protective ground (frame ground or chassis ground) AA GWG, FG, or CG 2 Transmit data (send data) BA TD, SD, TxD 3 Receive data BB RD, RxD 4 Request to send CA RS, RTS 5 Clear to send CB CS, CTS 6 Data set ready (modem ready) CC DSR, MR 7 Signal ground (reference ground) AB SG, GND 8 Receive line signal detect (carrier detect or data CF RLSD, CD, carrier detect) DCD 9 Unassigned — — 10 Unassigned — — 11 Unassigned — — 12 Secondary receive line signal detect (secondary carrier SCF SRLSD, detect or secondary data carrier detect) SCD, SDCD 13 Secondary clear to send SCB SCS, SCTS 14 Secondary transmit data (secondary send data) SBA STD, SSD, STxD 15 Transmit signal element timing—DCE (serial clock DB TSET, transmit—DCE) SCT-DCE 16 Secondary receive data SBB SRD, SRxD 17 Receive signal element timing (serial clock receive) DD RSET, SCR 18 Unassigned — — 19 Secondary request to send SCA SRS, SRTS 20 Data terminal ready CD DTR 21 Signal quality detect CG SQD 22 Ring indicator CE RI 23 Data signal rate selector CH DSRS 24 Transmit signal element timing—DTE (serial clock DA TSET, transmit—DTE) SCT-DTE 25 Unassigned — — The functions of the 25 RS-232 pins are summarized here for a DTE interfacing with a DCE where the DCE is a data communications modem: Pin 1—protective ground, frame ground, or chassis ground (GWG, FG, or CG). Pin 1 is connected to the chassis and used for protection against accidental electrical shock. Pin 1 is generally connected to signal ground (pin 7). Pin 2—transmit data or send data (TD, SD, or TxD). Serial data on the primary data channel are transported from the DTE to the DCE on pin 2. Primary data are the ac- tual source information transported over the interface. The transmit data line is a transmit line for the DTE but a receive line for the DCE. The DTE may hold the TD line at a logic 1 voltage level when no data are being transmitted and between char- acters when asynchronous data are being transmitted. Otherwise, the TD driver is en- abled by an active condition on pin 5 (clear to send). Pin 3—receive data (RD or RxD). Pin 3 is the second primary data pin. Serial data are transported from the DCE to the DTE on pin 3. Pin 3 is the receive data pin for the DTE and the transmit data pin for the DCE. The DCE may hold the TD line at a logic 1 voltage level when no data are being transmitted or when pin 8 (RLSD) is in- active. Otherwise, the RD driver is enabled by an active condition on pin 8. Pin 4—request to send (RS or RTS). For half-duplex data transmission, the DTE uses pin 4 to request permission from the DCE to transmit data on the primary data channel. When the DCE is a modem, an active condition on RTS turns on the modem’s analog Fundamental Concepts of Data Communications 190
- 196. Fundamental Concepts of Data Communications carrier.TheRTSandCTSsignalsareusedtogethertocoordinatehalf-duplexdatatrans- mission between the DTE and DCE. For full-duplex data transmission, RTS can be held active permanently. The RTS driver is enabled by an active condition on pin 6 (data set ready). Pin 5—clear to send (CS or CTS). The CTS signal is a handshake from the DCE to the DTE (i.e., modem to LCU) in response to an active condition on RTS. An active condition on CTS enables the TD driver in the DTE. There is a predetermined time delay between when the DCE receives an active condition on the RTS signal and when the DCE responds with an active condition on the CTS signal. Pin 6—data set ready or modem ready (DSR or MR). DSR is a signal sent from the DCE to the DTE to indicate the availability of the communications channel. DSR is active only when the DCE and the communications channel are available. Under nor- mal operation, the modem and the communications channel are always available. However, there are five situations when the modem or the communications channel are not available: 1. The modem is shut off (i.e., has no power). 2. The modem is disconnected from the communications line so that the line can be used for normal telephone voice traffic (i.e., in the voice rather than the data mode). 3. The modem is in one of the self-test modes (i.e., analog or digital loopback). 4. The telephone company is testing the communications channel. 5. On dial-up circuits, DSR is held inactive while the telephone switching system is establishing a call and when the modem is transmitting a specific response (an- swer) signal to the calling station’s modem. An active condition on the DSR lead enables the request to send driver in the DTE, thus giving the DSR lead the highest priority of the RS-232 control leads. Pin 7—signal ground or reference ground (SG or GND). Pin 7 is the signal reference (return line) for all data, control, and timing signals (i.e., all pins except pin 1, chas- sis ground). Pin 8—receive line signal detect, carrier detect, or data carrier detect (RLSD, CD, or DCD). The DCE uses this pin to signal the DTE when it determines that it is receiv- ing a valid analog carrier (data carrier). An active RLSD signal enables the RD termi- nator in the DTE, allowing it to accept data from the DCE. An inactive RLSD signal disables the terminator for the DTE’s receive data pin, preventing it from accepting in- valid data. On half-duplex data circuits, RLSD is held inactive whenever RTS is active. Pin 9—unassigned. Pin 9 is non–EIA specified; however, it is often held at 12 Vdc for test purposes (P). Pin 10—unassigned. Pin 10 is non–EIA specified; however, it is often held at 12Vdc for test purposes (P). Pin 11—unassigned. Pin 11 is non–EIA specified; however, it is often designated as equalizer mode (EM) and used by the modem to signal the DTE when the modem is self-adjusting its internal equalizers because error performance is suspected to be poor. When the carrier detect signal is active and the circuit is inactive, the modem is retraining (resynchronizing), and the probability of error is high. When the receive line signal detect (pin 8) is active and EM is inactive, the modem is trained, and the probability of error is low. Pin 12—secondary receive line signal detect, secondary carrier detect, or secondary data carrier detect (SRLSD, SCD, or SDCD). Pin 12 is the same as RLSD (pin 8), except for the secondary data channel. Pin 13—secondary clear to send. The SCTS signal is sent from DCE to the DTE as a response (handshake) to the secondary request to send signal (pin 19). 191
- 197. Fundamental Concepts of Data Communications Pin 14—secondary transmit data or secondary send data (STD, STD, or STxD). Di- agnostic data are transmitted from the DTE to the DCE on this pin. STD is enabled by an active condition on SCTS. Pin 15—transmission signal element timing or (serial clock transmit) DCE (TSET, SCT-DCE). With synchronous modems, the transmit clocking signal is sent from the DCE to the DTE on this pin. Pin 16—secondary received data (SRD or SRxD). Diagnostic data are transmitted from the DCE to the DTE on this pin. The SRD driver is enabled by an active condi- tion on secondary receive line signal detect (SRLSD). Pin 17—receiver signal element timing or serial clock receive (RSET or SCR). When synchronous modems are used, clocking information recovered by the DCE is sent to the DTE on this pin. The receive clock is used to clock data out of the DCE and into the DTE on the receive data line. The clock frequency is equal to the bit rate on the primary data channel. Pin 18—unassigned. Pin 11 is non–EIA specified; however, it is often used for the local loopback (LL) signal. Local loopback is a control signal sent from the DTE to the DCE placing the DCE (modem) into an analog loopback condition. Analog and digital loopbacks are described in a later section of this chapter. Pin 19—secondary request to send (SRS or SRTS). SRTS is used by the DTE to bid for the secondary data channel from the DCE. SRTS and SCTS coordinate the flow of data on the secondary data channel. Pin 20—data terminal ready (DTR). The DTE sends signals to the DCE on the DTR line concerning the availability of the data terminal equipment. DTR is used pri- marily with dial-up circuits to handshake with ring indicator (pin 22). The DTE dis- ables DTR when it is unavailable, thus instructing the DCE not to answer an in- coming call. Pin 21—signal quality detector (SQD). The DCE sends signals to the DTE on this line indicating the quality of the received analog carrier. An inactive (low) signal on SQD tells the DTE that the incoming signal is marginal and that there is a high like- lihood that errors are occurring. Pin 22—ring indicator (RI). The RI line is used primarily on dial-up data circuits for the DCE to inform the DTE that there is an incoming call. If the DTE is ready to re- ceive data, it responds to an active condition on RI with an active condition on DTR. DTR is a handshaking signal in response to an active condition on RI. Pin 23—data signal rate selector (DSRS). The DTE used this line to select one of two transmission bit rates when the DCE is equipped to offer two rates. (The data rate selector line can be used to change the transmit clock frequency.) Pin 24—transmit signal element timing or serial clock transmit–DTE (TSET, SCT- DTE). When synchronous modems are used, the transmit clocking signal is sent from the DTE to the DCE on this pin. Pin 24 is used only when the master clock is located in the DTE. Pin 25—unassigned. Pin 5 is non–EIA specified; however, it is sometimes used as a control signal from the DCE to the DTE to indicate that the DCE is in either the re- mote or local loopback mode. For asynchronous transmission using either the DB9P/S-modular connector, only the following nine pins are provided: 1. Receive line signal detect 2. Receive data 192
- 198. Fundamental Concepts of Data Communications Character time 10 ms time (ms) 0 Binary data +10 V –10 V 0 V 1 0 0 0 0 0 1 1 0 1 2 3 4 5 6 7 8 10 9 Start bit b0 LSB b1 b2 b3 b4 b5 b6 MSB Stop bit 1 ASCII Character Parity bit (even) 1 ms tb 1st bit transmitted last bit transmitted FIGURE 24 RS-232 data timing diagram—ASCII upper case letter A, 1 start bit, even parity, and one stop bit 3. Transmit data 4. Data terminal ready 5. Signal ground 6. Data set ready 7. Request to send 8. Clear to send 9. Ring indicator 11-1-3 RS-232 signals. Figure 24 shows the timing diagram for the transmission of one asynchronous data character over the RS-232 interface. The character is comprised of one start bit, one stop bit, seven ASCII character bits, and one even-parity bit. The trans- mission rate is 1000 bps, and the voltage level for a logic 1 is 10 V and for a logic 0 is 10 V. The time of one bit is 1 ms; therefore, the total time to transmit one ASCII charac- ter is 10 ms. 11-1-4 RS-232. Asynchronous Data Transmission. Figures 25a and b show the functional block diagram for the drivers and terminators necessary for transmission of asynchronous data over the RS-232 interface between a DTE and a DCE that is a mo- dem. As shown in the figure, only the first eight pins of the interface are required, which includes the following signals: signal ground and chassis ground, transmit data and re- ceive data, request to send, clear to send, data set ready, and receive line signal detect. Figure 26a shows the transmitter timing diagram for control and data signals for a typical asynchronous data transmission over an RS-232 interface with the following pa- rameters: Modem RTS-CTS delay 50 ms DTE primary data message length 100 ms 193
- 199. Fundamental Concepts of Data Communications DTE Pin 1 (chassis ground) Pin 2 (TD) Pin 4 (RTS) Pin 5 (CTS) Pin 6 (DSR) Drivers Terminators Pin 7 (signal ground) RS-232 Interface (a) DCE (Modem) RTS/CTS delay DTE Pin 3 (RD) Pin 8 (RLSD) RS-232 Interface (b) DCE (Modem) Data demodulator Analog carrier detect circuit FIGURE 25 Functional block diagram for the drivers and terminators necessary for transmission of asynchronous data over the RS-232 interface between a DTE and a DCE (modem): (a) transmit circuits; (b) receive circuits Modem training sequence 50 ms Propagation time 10 ms Modem RLSD turn-off delay time 5 ms When the DTE wishes to transmit data on the primary channel, it enables request to send (t 0).After a predetermined RTS/CTS time delay time, which is determined by the modem (50 ms for this example), CTS goes active. During the 50-ms RTS/CTS delay, the modem outputs an analog carrier that is modulated by a unique bit pattern called a training sequence. The training sequence for asynchronous modems is generally nothing more than a series of logic 1s that produce 50 ms of continuous mark frequency. The analog carrier is used to ini- tialize the communications channel and the distant receive modem (with synchronous modems, the training sequence is more involved, as it would also synchronize the carrier and clock recovery circuits in the distant modem). After the RTS/CTS delay, the transmit data (TD) line is enabled, and the DTE begins transmitting user data. When the transmission is 194
- 200. Fundamental Concepts of Data Communications t = 100 ms Transmit station (a) DSR RTS 150 ms 100 ms User data 100 ms Digitally modulated carrier (transmitted user data) 100 ms 50 ms RTS/CTS delay 50 ms 50 ms Digitally modulated carrier (training sequence) CTS TD Analog carrier t = 150 ms t = 200 ms t = 50 ms t = 0 ms FIGURE 26 Typical timing diagram for control and data signals for asynchronous data trans- mission over the RS-232 interface between a DTE and a DCE (modem): (a) transmit timing diagram; (b) receive timing diagram Receive station t = 10 ms t = 60 ms Pd = 10 ms t = 0 ms 50 ms RD RLSD RLSD turn-off delay = 10 ms 150 ms t = 160 ms t = 170 ms Digitally modulated carrier (demodulated user data) 100 ms Digitally modulated carrier (training sequence) (b) 195
- 201. Fundamental Concepts of Data Communications Table 10a RS-449 Pin Designations (37-Pin Connector) Pin Number Pin Name EIA Nomenclature 1 Shield None 19 Signal SG 37 Send common SC 20 Receive common RC 28 Terminal in service IS 15 Incoming call IC 12, 30 Terminal ready TR 11, 29 Data mode DM 4, 22 Send data SD 6, 24 Receive data RD 17, 35 Terminal timing TT 5, 23 Send timing ST 8, 26 Receive timing RT 7, 25 Request to send RS 9, 27 Clear to send CS 13, 31 Receiver ready RR 33 Signal quality SQ 34 New signal NS 16 Select frequency SF 2 Signal rate indicator SI 10 Local loopback LL 14 Remote loopback RL 18 Test mode TM 32 Select standby SS 36 Standby indicator SB complete (t 150 ms), RTS goes low, which turns off the modem’s analog carrier. The mo- dem acknowledges the inactive condition of RTS with an inactive condition on CTS. At the distant end (see Figure 26b), the receive modem receives a valid analog car- rier after a 10-ms propagation delay (Pd) and enables RLSD. The DCE sends an active RLSD signal across the RS-232 interface cable to the DT, which enables the receive data line (RD). However, the first 50 ms of the receive data is the training sequence, which is ig- nored by the DTE, as it is simply a continuous stream of logic 1s. The DTE identifies the beginning of the user data by recognizing the high-to-low transition caused by the first start bit (t 60 ms). At the end of the message, the DCE holds RLSD active for a predetermined RLSD turn-off delay time (10 ms) to ensure that all the data received have been demodu- lated and outputted onto the RS-232 interface. 11-2 RS-449 Serial Interface Standards In the mid-1970s, it appeared that data rates had exceeded the capabilities of the RS-232 in- terface. Consequently, in 1977, the EIA introduced the RS-449 serial interface with the in- tention of replacing the RS-232 interface. The RS-449 interface specifies a 37-pin primary connector (DB37) and a nine-pin secondary connector (DB9) for a total of 46 pins, which pro- vide more functions, faster data transmission rates, and spans greater distances than the RS- 232 interface.The RS-449 is essentially an updated version of the RS-232 interface except the RS-449 standard outlines only the mechanical and functional specifications of the interface. The RS-449 primary cable is for serial data transmission, while the secondary cable is for diagnostic information. Table 10a lists the 37 pins of the RS-449 primary cable and their designations, and Table 10b lists the nine pins of the diagnostic cable and their desig- nations. Note that the acronyms used with the RS-449 interface are more descriptive than those recommended by the EIA for the RS-232 interface. The functions specified by the RS-449 are very similar to the functions specified by the RS-232. The major difference 196
- 202. Fundamental Concepts of Data Communications Table 10b RS-449 Pin Designations (Nine-Pin Connector) Pin Number Pin Name EIA Nomenclature 1 Shield None 5 Signal ground SG 9 Send common SC 2 Receive common RC 3 Secondary send data SSD 4 Secondary receive data SRD 7 Secondary request to send SRS 8 Secondary clear to send SCS 6 Secondary receiver ready SRF between the two standards is the separation of the primary data and secondary diagnostic channels onto two separate cables. The electrical specifications for the RS-449 were specified by the EIA in 1978 as ei- ther the RS-422 or the RS-423 standard. The RS-449 standard, when combined with RS- 422A or RS-423A, were intended to replace the RS-232 interface. The primary goals of the new specifications are listed here: 1. Compatibility with the RS-232 interface standard 2. Replace the set of circuit names and mnemonics used with the RS-232 interface with more meaningful and descriptive names 3. Provide separate cables and connectors for the primary and secondary data channels 4. Provide single-ended or balanced transmission 5. Reduce crosstalk between signal wires 6. Offer higher data transmission rates 7. Offer longer distances over twisted-pair cable 8. Provide loopback capabilities 9. Improve performance and reliability 10. Specify a standard connector The RS-422A standard specifies a balanced interface cable capable of operating up to 10 Mbps and span distances up to 1200 meters. However, this does not mean that 10 Mbps can be transmitted 1200 meters.At 10 Mbps, the maximum distance is approximately 15 meters, and 90 kbps is the maximum bit rate that can be transmitted 1200 meters. The RS-423A standard specifies an unbalanced interface cable capable of operating at a maximum transmission rate of 100 kbps and span a maximum distance of 90 meters. The RS-442A and RS-443A standards are similar to ITU-T V.11 and V.10, respectively. With a bidirectional unbalanced line, one wire is at ground potential, and the currents in the two wires may be different. With an unbalanced line, interference is induced into only one signal path and, therefore, does not cancel in the terminator. The primary objective of establishing the RS-449 interface standard was to maintain compatibility with the RS-232 interface standard. To achieve this goal, the EIA divided the RS-449 into two categories: category I and category II circuits. Category I circuits include only circuits that are compatible with the RS-232 standard. The remaining circuits are clas- sified as category II. Category I and category II circuits are listed in Table 11. Category I circuits can function with either the RS-422A (balanced) or the RS-423A (unbalanced) specifications. Category I circuits are allotted two adjacent wires for each RS- 232-compatible signal, which facilitates either balanced or unbalanced operation. Category II circuits are assigned only one wire and, therefore, can facilitate only unbalanced (RS- 423A) specifications. 197
- 203. Fundamental Concepts of Data Communications Table 11 RS-449 Category I and Category II Circuits Category I SD Send data (4, 22) RD Receive data (6, 24) TT Terminal timing (17, 35) ST Send timing (5, 23) RT Receive timing (8, 26) RS Request to send (7, 25) CS Clear to send (9, 27) RR Receiver ready (13, 31) TR Terminal ready (12, 30) DM Data mode (11, 29) Category II SC Send common (37) RC Receive common (20) IS Terminal in service (28) NS New signal (34) SF Select frequency (16) LL Local loopback (10) RL Remote loopback (14) TM Test mode (18) SS Select standby (32) SB Standby indicator (36) The RS-449 interface provides 10 circuits not specified in the RS-232 standard: 1. Local loopback (LL, pin 10). Used by the DTE to request a local (analog) loop- back from the DCE 2. Remote loopback (RL, pin 14). Used by the DTE to request a remote (digital) loopback from the distant DCE 3. Select frequency (SF, pin 16). Allows the DTE to select the DCE’s transmit and receive frequencies 4. Test mode (TM, pin 18). Used by the DTE to signal the DCE that a test is in progress 5. Receive common (RC, pin 20). Common return wire for unbalanced signals prop- agating from the DCE to the DTE 6. Terminal in service (IS, pin 28). Used by the DTE to signal the DCE whether it is operational 7. Select standby (SS, pin 32). Used by the DTE to request that the DCE switch to standby equipment in the event of a failure on the primary equipment 8. New signal (NS, pin 34). Used with a modem at the primary location of a multi- point data circuit so that the primary can resynchronize to whichever secondary is transmitting at the time 9. Standby indicator (SB, pin 36). Intended to be by the DCE as a response to the SS signal to notify the DTE that standby equipment has replaced the primary equipment 10. Send common (SC, pin 37). Common return wire for unbalanced signals propa- gating from the DTE to the DCE 11-3 RS-530 Serial Interface Standards Since the data communications industry did not readily adopt the RS-449 interface, it came and went virtually unnoticed by most of industry. Consequently, in 1987 the EIA introduced another new standard, the RS-530 serial interface, which was intended to operate at data rates 198
- 204. Fundamental Concepts of Data Communications Table 12 RS-530 Pin Designations Signal Name Pin Number(s) Shield 1 Transmit dataa 2, 14 Receive dataa 3, 16 Request to senda 4, 19 Clear to senda 5, 13 DCE readya 6, 22 DTE readya 20, 23 Signal ground 7 Receive line signal detecta 8, 10 Transmit signal element timing (DCE source)a 15, 12 Receive signal element timing (DCE source)a 17, 9 Local loopbackb 18 Remote loopbackb 21 Transmit signal element timing (DTE source)a 24, 11 Test modeb 25 a Category I circuits (RS-422A). b Category II circuits (RS-423A). between 20 kbps and 2 Mbps using the same 25-pin DB-25 connector used by the RS-232 interface. The pin functions of the RS-530 interface are essentially the same as the RS-449 category I pins with the addition of three category II pins: local loopback, remote loopback, and test mode. Table 12 lists the 25 pins for the RS-530 interface and their designations. Like the RS-449 standard, the RS-530 interface standard does not specify electrical pa- rameters. The electrical specifications for the RS-530 are outlined by either the RS-422A or the RS-423A standard. The RS-232, RS-449, and RS-530 interface standards provide speci- ficationsforansweringcalls,butdonotprovidespecificationsforinitiatingcalls(i.e.,dialing). The EIA has a different standard, RS-366, for automatic calling units. The principal use of the RS-366 is for dial backup of private-line data circuits and for automatic dialing of remote ter- minals. 12 DATA COMMUNICATIONS MODEMS The most common type of data communications equipment (DEC) is the data communica- tions modem. Alternate names include datasets, dataphones, or simply modems. The word modem is a contraction derived from the words modulator and demodulator. In the 1960s, the business world recognized a rapidly increasing need to exchange digi- talinformationbetweencomputers,computerterminals,andothercomputer-controlledequip- ment separated by substantial distances. The only transmission facilities available at the time were analog voice-band telephone circuits. Telephone circuits were designed for transporting analogvoicesignalswithinabandwidthofapproximately300Hzto3000Hz.Inaddition,tele- phone circuits often included amplifiers and other analog devices that could not propagate dig- ital signals. Therefore, voice-band data modems were designed to communicate with each other using analog signals that occupied the same bandwidth used for standard voice telephone communications. Data communications modems designed to operate over the limited band- width of the public telephone network are called voice-band modems. Because digital information cannot be transported directly over analog transmission media (at least not in digital form), the primary purpose of a data communications modem is to interface computers, computer networks, and other digital terminal equipment to analog communications facilities. Modems are also used when computers are too far apart to be 199
- 205. Fundamental Concepts of Data Communications POTS analog channel Digitally-modulated analog signals Digital pulses DCE (Modem) DCE (Modem) DTE (PC) DTE (PC) Serial Interface (RS-232) Serial Interface (RS-232) Digital pulses FIGURE 27 Data communications modems - POTS analog channel directly interconnected using standard computer cables. In the transmitter (modulator) sec- tion of a modem, digital signals are encoded onto an analog carrier. The digital signals mod- ulate the carrier, producing digitally modulated analog signals that are capable of being trans- ported through the analog communications media. Therefore, the output of a modem is an analog signal that is carrying digital information. In the receiver section of a modem, digitally modulated analog signals are demodulated. Demodulation is the reverse process of modula- tion. Therefore, modem receivers (demodulators) simply extract digital information from dig- itally modulated analog carriers. The most common (and simplest) modems available are ones intended to be used to interface DTEs through a serial interface to standard voice-band telephone lines and pro- vide reliable data transmission rates from 300 bps to 56 kbps. These types of modems are sometimes called telephone-loop modems or POTS modems, as they are connected to the telephone company through the same local loops that are used for voice telephone circuits. More sophisticated modems (sometimes called broadband modems) are also available that are capable of transporting data at much higher bit rates over wideband communications channels, such as those available with optical fiber, coaxial cable, microwave radio, and satellite communications systems. Broadband modems can operate using a different set of standards and protocols than telephone loop modems. A modem is, in essence, a transparent repeater that converts electrical signals received in digital form to electrical signals in analog form and vice versa. A modem is transparent, as it does not interpret or change the information contained in the data. It is a repeater, as it is not a destination for data—it simply repeats or retransmits data. A modem is physically located between digital terminal equipment (DTE) and the analog communications chan- nel. Modems work in pairs with one located at each end of a data communications circuit. The two modems do not need to be manufactured by the same company; however, they must use compatible modulation schemes, data encoding formats, and transmission rates. Figure 27 shows how a typical modem is used to facilitate the transmission of digital data between DTEs over a POTS telephone circuit.At the transmit end, a modem receives dis- crete digital pulses (which are usually in binary form) from a DTE through a serial digital in- terface (such as the RS-232). The DCE converts the digital pulses to analog signals. In essence, a modem transmitter is a digital-to-analog converter (DAC). The analog signals are then outputted onto an analog communications channel where they are transported through the system to a distant receiver. The equalizers and bandpass filters shape and band-limit the signal.At the destination end of a data communications system, a modem receives analog sig- nals from the communications channel and converts them to digital pulses. In essence, a mo- dem receiver is an analog-to-digital converter (ADC). The demodulated digital pulses are then outputted onto a serial digital interface and transported to the DTE. 200
- 206. Fundamental Concepts of Data Communications 12-1 Bits per Second versus Baud The parameters bits per second (bps) and baud are often misunderstood and, consequently, misused. Baud, like bit rate, is a rate of change; however, baud refers to the rate of change of the signal on the transmission medium after encoding and modulation have occurred. Bit rate refers to the rate of change of a digital information signal, which is usually binary. Baud is the reciprocal of the time of one output signaling element, and a signaling element may represent several information bits. A signaling element is sometimes called a symbol and could be encoded as a change in the amplitude, frequency, or phase. For example, binary signals are generally encoded and transmitted one bit at a time in the form of discrete volt- age levels representing logic 1s (highs) and logic 0s (lows). A baud is also transmitted one at a time; however, a baud may represent more than one information bit. Thus, the baud of a data communications system may be considerably less than the bit rate. 12-2 Bell System–Compatible Modems At one time, Bell System modems were virtually the only modems in existence. This is be- causeATT operating companies once owned 90% of the telephone companies in the United States, and theATT operating tariff allowed only equipment manufactured byWestern Elec- tric Company (WECO) and furnished by Bell System operating companies to be connected to ATT telephone lines. However, in 1968,ATT lost a landmark Supreme Court decision, the Carterfone decision, which allowed equipment manufactured by non-Bell companies to in- terconnect to the vastATT communications network, provided that the equipment met Bell System specifications.The Carterfone decision began theinterconnect industry, which has led to competitive data communications offerings by a large number of independent companies. The operating parameters for Bell System modems are the models from which the in- ternational standards specified by the ITU-T evolved. Bell System modem specifications apply only to modems that existed in 1968; therefore, their specifications pertain only to modems operating at data transmission rate of 9600 bps or less. Table 11 summarizes the parameters for Bell System–equivalent modems. 12-3 Modem Block Diagram Figure 28 shows a simplified block diagram for a data communications modem. For sim- plicity, only the primary functional blocks of the transmitter and receiver are shown. The Modulator and carrier circuit Transmitter (D/A converter) Receiver (A/D converter) Analog signals Analog signals Digital pulses (data) Digital pulses (data) Bandpass filter and pre-equalizer Bandpass filter and post-equalizer Serial interface circuit DTE Telco interface Telco local loop Demodulator circuit Serial interface 2-wire or 4-wire FIGURE 28 Simplified block diagram for an asynchronous FSK modem 201
- 207. Fundamental Concepts of Data Communications basic principle behind a modem transmitter is to convert information received from the DTE in the form of binary digits (bits) to digitally modulated analog signals. The reverse process is accomplished in the modem receiver. The primary blocks of a modem are described here: 1. Serial interface circuit. Interfaces the modem transmitter and receiver to the serial in- terface. The transmit section accepts digital information from the serial interface, converts it totheappropriatevoltagelevels,andthendirectstheinformationtothemodulator.Thereceive section receives digital information from the demodulator circuit, converts it to the appropri- ate voltage levels, and then directs the information to the serial interface. In addition, the serial interfacecircuitmanagestheflowofcontrol,timing,anddatainformationtransferredbetween the DTE and the modem, which includes handshaking signals and clocking information. 2. Modulator circuit. Receives digital information from the serial interface circuit. The digital information modulates an analog carrier, producing a digitally modulated ana- log signal. In essence, the modulator converts digital changes in the information to analog changes in the carrier. The output from the modulator is directed to the transmit bandpass filter and equalizer circuit. 3. Bandpass filter and equalizer circuit. There are bandpass filter and equalizer cir- cuits in both the transmitter and receiver sections of the modem. The transmit bandpass fil- ter limits the bandwidth of the digitally modulated analog signals to a bandwidth appropri- ate for transmission over a standard telephone circuit. The receive bandpass filter limits the bandwidth of the signals allowed to reach the demodulator circuit, thus reducing noise and improving system performance. Equalizer circuits compensate for bandwidth and gain im- perfections typically experienced on voiceband telephone lines. 4. Telco interface circuit. The primary functions of the telco interface circuit are to match the impedance of the modem to the impedance of the telephone line and regulate the amplitude of the transmit signal. The interface also provides electrical isolation and pro- tection and serves as the demarcation (separation) point between subscriber equipment and telephone company–provided equipment. The telco line can be two-wire or four-wire, and the modem can operate half or full duplex. When the telephone line is two wire, the telco interface circuit would have to perform four-wire-to-two-wire and two-wire-to-four-wire conversions. 5. Demodulator circuit. Receives modulated signals from the bandpass filter and equalizer circuit and converts the digitally modulated analog signals to digital signals. The output from the demodulator is directed to the serial interface circuit, where it is passed on to the serial interface. 6. Carrier and clock generation circuit. The carrier generation circuit produces the analog carriers necessary for the modulation and demodulation processes. The clock gen- eration circuit generates the appropriate clock and timing signals required for performing transmit and receive functions in an orderly and timely fashion. 12-4 Modem Classifications Data communications modems can be generally classified as either asynchronous or syn- chronous and use one of the following digital modulation schemes: amplitude-shift key- ing (ASK), frequency-shift keying (FSK), phase-shift keying (PSK), or quadrature ampli- tude modulation (QAM). However, there are several additional ways modems can be classified, depending on which features or capabilities you are trying to distinguish. For example, modems can be categorized as internal or external; low speed, medium speed, high speed, or very high speed; wide band or voice band; and personal or commercial. Re- gardless of how modems are classified, they all share a common goal, namely, to convert digital pulses to analog signals in the transmitter and analog signals to digital pulses in the receiver. 202
- 208. Fundamental Concepts of Data Communications Some of the common features provided data communications modems are listed here: 1. Automatic dialing, answering, and redialing 2. Error control (detection and correction) 3. Caller ID recognition 4. Self-test capabilities, including analog and digital loopback tests 5. Fax capabilities (transmit and receive) 6. Data compression and expansion 7. Telephone directory (telephone number storage) 8. Adaptive transmit and receive data transmission rates (300 bps to 56 kbps) 9. Automatic equalization 10. Synchronous or asynchronous operation 12-5 Asynchronous Voice-Band Modems Asynchronous modems can be generally classified as low-speed voice-band modems, as they are typically used to transport asynchronous data (i.e., data framed with start and stop bits). Synchronous data are sometimes used with an asynchronous modem; however, it is not particularly practical or economical. Synchronous data transported by asynchronous modems is called isochronous transmission. Asynchronous modems use relatively simple modulation schemes, such as ASK or FSK, and are restricted to relatively low-speed ap- plications (generally less than 2400 bps), such as telemetry and caller ID. There are several standard asynchronous modems designed for low-speed data appli- cations using the switched public telephone network. To operate full duplex with a two-wire dial-up circuit, it is necessary to divide the usable bandwidth of a voice-band circuit in half, creating two equal-capacity data channels. A popular modem that does this is the Bell Sys- tem 103–compatible modem. 12-5-1 Bell system 103–compatible modem. The 103 modem is capable of full-du- plex operation over a two-wire telephone line at bit rates up to 300 bps. With the 103 mo- dem, there are two data channels, each with their own mark and space frequencies. One data channel is called the low-band channel and occupies a bandwidth from 300 Hz to 1650 Hz (i.e., the lower half of the usable voice band). A second data channel, called the high-band channel, occupies a bandwidth from 1650 Hz to 3000 Hz (i.e., the upper half of the usable voice band). The mark and space frequencies for the low-band channel are 1270 Hz and 1070 Hz, respectively. The mark and space frequencies for the high-band channel are 2225 Hz and 2025 Hz, respectively. Separating the usable bandwidth into two narrower bands is called frequency-division multiplexing (FDM). FDM allows full-duplex (FDX) transmission over a two-wire circuit, as signals can propagate in both directions at the same time without interfering with each other because the frequencies for the two directions of propagation are different. FDM allows full-duplex operation over a two-wire telephone cir- cuit. Because FDM reduces the effective bandwidth in each direction, it also reduces the maximum data transmission rates. A 103 modem operates at 300 baud and is capable of si- multaneous transmission and reception of 300 bps. 12-5-2 Bell system 202T/S modem. The 202T and 202S modem are identical ex- cept the 202T modem specifies four-wire, full-duplex operation, and the 202S modem spec- ifies two-wire, half-duplex operation. Therefore, the 202T is utilized on four-wire pri- vate-line data circuits, and the 202S modem is designed for the two-wire switched public telephone network. Probably the most common application of the 202 modem today is caller ID, which is a simplex system with the transmitter in the telephone office and the re- ceiver at the subscriber’s location. The 202 modem is an asynchronous 1200-baud trans- ceiver utilizing FSK with a transmission bit rate of 1200 bps over a standard voice-grade telephone line. 203
- 209. Fundamental Concepts of Data Communications 12-6 Synchronous Voice-Band Modems Synchronous modems use PSK or quadrature amplitude modulation (QAM) to transport syn- chronous data (i.e., data preceded by unique SYN characters) at transmission rates between 2400 bps and 56,000 bps over standard voice-grade telephone lines. The modulated carrier is transmitted to the distant modem, where a coherent carrier is recovered and used to demod- ulate the data. The transmit clock is recovered from the data and used to clock the received data into the DTE. Because of the addition of clock and carrier recovery circuits, synchro- nous modems are more complicated and, thus, more expensive than asynchronous modems. PSK is commonly used in medium speed synchronous voice-band modems, typically operating between 2400 bps and 4800 bps. More specifically, QPSK is generally used with 2400-bps modems and 8-PSK with 4800-bps modems. QPSK has a bandwidth efficiency of 2 bps/Hz; therefore, the baud rate and minimum bandwidth for a 2400-bps synchronous mo- dem are 1200 baud and 1200 Hz, respectively. The standard 2400-bps synchronous modem is the Bell System 201C or equivalent. The 201C modem uses a 1600-Hz carrier frequency and has an output spectrum that extends from approximately 1000 Hz to 2200 Hz. Because 8- PSK has a bandwidth efficiency of 3 bps/Hz, the baud rate and minimum bandwidth for 4800- bps synchronous modems are 1600 baud and 1600 Hz, respectively. The standard 4800-bps synchronous modem is the Bell System 208A. The 208A modem also uses a 1600-Hz carrier frequency but has an output spectrum that extends from approximately 800 Hz to 2400 Hz. Both the 201C and the 208A are full-duplex modems designed to be used with four-wire private-line circuits. The 201C and 208A modems can operate over two-wire dial-up circuits but only in the simplex mode. There are also half-duplex two-wire versions of both modems: the 201B and 208B. High-speed synchronous voice-band modems operate at 9600 bps and use 16-QAM modulation. 16-QAM has a bandwidth efficiency of 4 bps/Hz; therefore, the baud and min- imum bandwidth for 9600-bps synchronous modems is 2400 baud and 2400 Hz, respec- tively. The standard 9600-bps modem is the Bell System 209A or equivalent. The 209A uses a 1650-Hz carrier frequency and has an output spectrum that extends from approxi- mately 450 Hz to 2850 Hz. The Bell System 209A is a four-wire synchronous voice-band modem designed to be used on full-duplex private-line circuits. The 209B is the two-wire version designed for half-duplex operation on dial-up circuits. Table 13 summarizes the Bell System voice-band modem specifications. The modems listed in the table are all relatively low speed by modern standards. Today, the Bell System–compatible modems are used primarily on relatively simple telemetry circuits, such as remote alarm systems and on metropolitan and wide-area private-line data net- works, such as those used by department stores to keep track of sales and inventory. The more advanced, higher-speed data modems are described in a later section of this chapter. 12-7 Modem Synchronization During the request-to-send/clear-to-send (RTS/CTS) delay, a transmit modem outputs a special, internally generated bit pattern called a training sequence. This bit pattern is used to synchronize (train) the receive modem at the distant end of the communications channel. Depending on the type of modulation, transmission bit rate, and modem complexity, the training sequence accomplishes one or more of the following functions: 1. Initializes the communications channel, which includes disabling echo and estab- lishing the gain of automatic gain control (AGC) devices 2. Verifies continuity (activates RLSD in the receive modem) 3. Initialize descrambler circuits in receive modem 4. Initialize automatic equalizers in receive modem 5. Synchronize the receive modem’s carrier to the transmit modem’s carrier 6. Synchronize the receive modem’s clock to the transmit modem’s clock 204
- 210. Fundamental Concepts of Data Communications Table 13 Bell System Modem Specifications Bell System Transmission Operating Circuit Synchronization Transmission Designation Facility Mode Arrangement Mode Modulation Rate 103 Dial-up FDM/FDX Two wire Asynchronous FSK 300 bps 113A/B Dial-up FDM/FDX Two wire Asynchronous FSK 300 bps 201B Dial-up HDX Two wire Synchronous QPSK 2400 bps 201C Private line FDX Four wire Synchronous QPSK 2400 bps 202S Dial-up HDX Two wire Asynchronous FSK 1200 bps 202T Private line FDX Four wire Asynchronous FSK 1800 bps 208A Private line FDX Four wire Synchronous 8-PSK 4800 bps 208B Dial-up HDX Two wire Synchronous 8-PSK 4800 bps 209A Private line FDX Four wire Synchronous 16-QAM 9600 bps 209B Dial-up HDX Two wire Synchronous 16-QAM 9600 bps 212A Dial up HDX Two wire Asynchronous FSK 600 bps 212B Private line FDX Four wire Synchronous QPSK 1200 bps Dial-up switched telephone network Private line dedicated circuit FDM frequency-division multiplexing HDX half duplex FDX full duplex FSK frequency-shift keying QPSK four-phase PSK 8-PSK eight-phase PSK 16-QAM 16-state QAM 12-8 Modem Equalizers Equalization is the compensation for phase delay distortion and amplitude distortion in- herently present on telephone communications channels. One form of equalization provided by the telephone company is C-type conditioning, which is available only on private-line cir- cuits. Additional equalization may be performed by the modems themselves. Compromise equalizers are located in the transmit section of a modem and provide preequalization— they shape the transmitted signal by altering its delay and gain characteristics before the signal reaches the telephone line. It is an attempt by the modem to compensate for impair- ments anticipated in the bandwidth parameters of the communications line. When a modem is installed, the compromise equalizers are manually adjusted to provide the best error per- formance. Typically, compromise equalizers affect the following: 1. Amplitude only 2. Delay only 3. Amplitude and delay 4. Neither amplitude nor delay Compromise equalizer settings may be applied to either the high- or low-voice-band fre- quencies or symmetrically to both at the same time. Once a compromise equalizer setting has been selected, it can be changed only manually. The setting that achieves the best error performance is dependent on the electrical length of the circuit and the type of facilities that comprise it (i.e., one or more of the following: twisted-pair cable, coaxial cable, optical fiber cable, microwave, digital T-carriers, and satellite). Adaptive equalizers are located in the receiver section of a modem, where they pro- vide postequalization to the received signals. Adaptive equalizers automatically adjust their gain and delay characteristics to compensate for phase and amplitude impairments encoun- tered on the communications channel. Adaptive equalizers may determine the quality of the received signal within its own circuitry, or they may acquire this information from the 205
- 211. Fundamental Concepts of Data Communications demodulator or descrambler circuits. Whatever the case, the adaptive equalizer may contin- uously vary its settings to achieve the best overall bandwidth characteristics for the circuit. 13 ITU-T MODEM RECOMMENDATIONS Since the late 1980s, the International Telecommunications Union (ITU-T, formerly CCITT), which is headquartered in Geneva, Switzerland, has developed transmission stan- dards for data modems outside the United States. The ITU-T specifications are known as the V-series, which include a number indicating the standard (V.21, V.23, and so on). Some- times the V-series is followed by the French word bis, meaning “second,” which indicates that the standard is a revision of an earlier standard. If the standard includes the French word terbo, meaning “third,” the bis standard also has been modified. Table 14 lists some of the ITU-T modem recommendations. 13-1 ITU-T Modem Recommendation V.29 The ITU-T V.29 specification is the first internationally accepted standard for a 9600-bps data transmission rate. The V.29 standard is intended to provide synchronous data transmis- sion over four-wire leased lines. V.29 uses 16-QAM modulation of a 1700-Hz carrier fre- quency. Data are clocked into the modem in groups of four bits called quadbits, resulting in a 2400-baud transmission rate. Occasionally,V.29-compatible modems are used in the half- duplex mode over two-wire switched telephone lines. Pseudo full-duplex operation can be achieved over the two-wire lines using a method called ping-pong.With ping-pong, data sent to the modem at each end of the circuit by their respective DTE are buffered and automati- cally exchanged over the data link by rapidly turning the carriers on and off in succession. Pseudo full-duplex operation over a two-wire line can also be accomplished using sta- tistical duplexing. Statistical duplexing utilizes a 300-bps reverse data channel. The reverse channel allows a data operator to enter keyboard data while simultaneously receiving a file from the distant modem. By monitoring the data buffers inside the modem, the direction of data transmission can be determined, and the high- and low-speed channels can be reversed. 13-2 ITU-T Modem Recommendation V.32 The ITU-T V.32 specification provides for a 9600-bps data transmission rate with true full- duplex operation over four-wire leased private lines or two-wire switched telephone lines. V.32 also provides for data rates of 2400 bps and 4800 bps. V.32 specifies QAM with a car- rier frequency of 1800 Hz. V.32 is similar to V.29, except with V.32 the advanced coding technique trellis encoding is specified. Trellis encoding produces a superior signal-to-noise ratio by dividing the incoming data stream into groups of five bits called quintbits (M-ary, where M 25 32). The constellation diagram for 32-state trellis encoding was developed by Dr. Ungerboeck at the IBM Zuerich Research Laboratory and combines coding and modulation to improve bit error performance. The basic idea behind trellis encoding is to introduce controlled redundancy, which reduces channel error rates by doubling the num- ber of signal points on the QAM constellation. The trellis encoding constellation used with V.32 is shown in Figure 29. Full-duplex operation over two-wire switched telephone lines is achieved withV.32 us- ingatechniquecalledechocancellation.Echocancellationinvolvesaddinganinvertedreplica of the transmitted signal to the received signal.This allows the data transmitted from each mo- dem to simultaneously use the same carrier frequency, modulation scheme, and bandwidth. 13-3 ITU-T Modem Recommendation V.32bis and V.32terbo ITU-T recommendation V.32bis was introduced in 1991 and created a new benchmark for the data modem industry by allowing transmission bit rates of 14.4 kbps over standard 206
- 212. Fundamental Concepts of Data Communications Table 14 ITU-T V-Series Modem Standards ITU-T Designation Specification V.1 Defines binary 0/1 data bits as space/mark line conditions V.2 Limits output power levels of modems used on telephone lines V.4 Sequence of bits within a transmitted character V.5 Standard synchronous signaling rates for dial-up telephone lines V.6 Standard synchronous signaling rates for private leased communications lines V.7 List of modem terminology in English, Spanish, and French V.10 Unbalanced high-speed electrical interface specifications (similar to RS-423) V.11 Balanced high-speed electrical interface specifications (similar to RS-422) V.13 Simulated carrier control for full-duplex modem operating in the half-duplex mode V.14 Asynchronous-to-synchronous conversion V.15 Acoustical couplers V.16 Electrocardiogram transmission over telephone lines V.17 Application-specific modulation scheme for Group III fax (provides two-wire, half-duplex trellis-coded transmission at 7.2 kbps, 9.6 kbps, 12 kbps, and 14.4 kbps) V.19 Low-speed parallel data transmission using DTMF modems V.20 Parallel data transmission modems V.21 0-to-300 bps full-duplex two-wire modems similar to Bell System 103 V.22 1200/600 bps full-duplex modems for switched or dedicated lines V.22bis 1200/2400 bps two-wire modems for switched or dedicated lines V.23 1200/75 bps modems (host transmits 1200 bps and terminal transmits 75 bps). V.23 also sup- ports 600 bps in the high channel speed. V.23 is similar to Bell System 202. V.23 is used in Europe to support some videotext applications. V.24 Known in the United States as RS-232. V.24 defines only the functions of the interface cir- cuits, whereas RS-232 also defines the electrical characteristics of the connectors. V.25 Automatic answering equipment and parallel automatic dialing similar to Bell System 801 (defines the 2100-Hz answer tone that modems send) V.25bis Serial automatic calling and answering—CCITT equivalent to the Hayes AT command set used in the United States V.26 2400-bps four-wire modems identical to Bell System 201 for four-wire leased lines V.26bis 2400/1200 bps half-duplex modems similar to Bell System 201 for two-wire switched lines V.26terbo 2400/1200 bps full-duplex modems for switched lines using echo canceling V.27 4800 bps four-wire modems for four-wire leased lines similar to Bell System 208 with man- ual equalization V.27bis 4800/2400 bps four-wire modems same as V.27 except with automatic equalization V.28 Electrical characteristics for V.24 V.29 9600-bps four-wire full-duplex modems similar to Bell System 209 for leased lines V.31 Older electrical characteristics rarely used today V.31bis V.31 using optocouplers V.32 9600/4800 bps full-duplex modems for switched or leased facilities V.32bis 4.8-kbps, 7.2-kbps, 9.6-kbps, 12-kbps, and 14.4-kbps modems and rapid rate regeneration for full-duplex leased lines V.32terbo Same as V.32bis except with the addition of adaptive speed leveling, which boosts transmis- sion rates to as high as 21.6 kbps V.33 12.2 kbps and 14.4 kbps for four-wire leased communications lines V.34 (V. fast) 28.8-kbps data rates without compression V.34 Enhanced specifications of V.34 V.35 48-kbps four-wire modems (no longer used) V.36 48-kbps four-wire full-duplex modems V.37 72-kbps four-wire full-duplex modems V.40 Method teletypes use to indicate parity errors V.41 An older obsolete error-control scheme V.42 Error-correcting procedures for modems using asynchronous-to-synchronous conversion (V.22, B.22bis, V.26terbo, V.32, and V.32bis, and LAP M protocol) V.42bis Lempel-Ziv-based data compression scheme used with V.42 LAP M V.50 Standard limits for transmission quality for modems V.51 Maintenance of international data circuits (Continued) 207
- 213. Fundamental Concepts of Data Communications 90° 11111 –90° 180° 0° 4 2 –2 –4 cos –cos –sin sin 01000 10010 00000 01111 01101 00011 00111 01001 01011 01110 01100 00100 10101 10011 10100 11001 –4 –2 2 4 11110 11010 11101 10000 10111 10001 11100 11011 10110 01010 11000 00101 00010 00110 00001 FIGURE 29 V.32 constellation diagram using Trellis encoding Table 14 (Continued) ITU-T Designation Specification V.52 Apparatus for measuring distortion and error rates for data transmission V.53 Impairment limits for data circuits V.54 Loop test devices of modems V.55 Impulse noise-measuring equipment V.56 Comparative testing of modems V.57 Comprehensive tests set for high-speed data transmission V.90 Asymmetrical data transmission—receive data rates up to 56 kbps but restricts transmission bit rates to 33.6 kbps V.92 Asymmetrical data transmission—receive data rates up to 56 kbps but restricts transmission bit rates to 48 kbps V.100 Interconnection between public data networks and public switched telephone networks V.110 ISDN terminal adaptation V.120 ISDN terminal adaptation with statistical multiplexing V.230 General data communications interface, ISO layer 1 voice-band telephone channels. V.32bis uses a 64-point signal constellation with each sig- naling condition representing six bits of data. The constellation diagram for V.32 is shown in Figure 30. The transmission bit rate for V.32 is six bits/code 2400 codes/second 14,400 bps. The signaling rate (baud) is 2400. V.32bis also includes automatic fall-forward and fall-back features that allow the mo- dem to change its transmission rate to accommodate changes in the quality of the commu- 208
- 214. Fundamental Concepts of Data Communications 90° 0000110 –90° 180° 0° cos –cos –sin sin 0000011 1100100 1011101 1010110 1010011 1000100 0001101 0001000 1110011 1001100 0000101 0000000 1100011 1011010 1010101 1010000 1000011 0001010 0000010 0001111 1001011 0000111 1000110 1101001 1110101 1111001 1100101 1011001 1101100 0011010 1011111 0100011 0011111 0101011 0010111 1101011 1010111 1100110 0010000 0100110 0101110 0100000 1101110 1000000 1001001 0010101 0101001 0110101 0111001 0100101 0011001 1000101 0001001 1111010 0101100 0111010 1111111 0110011 0111111 0111011 0110111 1111011 1110111 1110110 0011000 0110110 0111110 0101000 1111110 1001000 1000001 0011101 0100001 0111101 0110001 0101101 0010001 1001101 0000001 1101010 0100100 0101010 0100010 0010100 1100010 0000100 1101111 0010011 0101111 0011011 0100111 1011011 1100111 1011000 0010110 0011110 1101000 1011110 1100001 1111101 1110001 1101101 1010001 1001010 1000010 1010100 1001111 0001011 1000111 0001110 1110100 1111000 0110100 0111000 0111100 0110010 0011100 1110010 0001100 0110000 1111100 0010010 1011100 1010010 1110000 1001110 1100000 FIGURE 30 V.33 signal constellation diagram using Trellis encoding nications line. The fall-back feature slowly reduces the transmission bit rate to 12.2 kbps, 9.6 kbps, or 4.8 kbps if the quality of the communications line degrades. The fall-forward feature gives the modem the ability to return to the higher transmission rate when the qual- ity of the communications channel improves. V.32bis support Group III fax, which is the transmission standard that outlines the connection procedures used between two fax ma- chines or fax modems. V.32bis also specifies the data compression procedure used during transmissions. In August 1993, U.S. Robotics introduced V.32terbo. V.32terbo includes all the fea- tures of V.32bis plus a proprietary technology called adaptive speed leveling. V.32terbo includes two categories of new features: increased data rates and enhanced fax abilities. V.32terbo also outlines the new 19.2-kbps data transmission rate developed by ATT. 13-4 ITU-T Modem Recommendation V.33 ITU-T specificationV.33 is intended for modems that operate over dedicated two-point, pri- vate-line four-wire circuits. V.33 uses trellis coding and is similar to V.32 except a V.33 sig- naling element includes six information bits and one redundant bit, resulting in a data trans- mission rate of 14.4 kbps, 2400 baud, and an 1800-Hz carrier. The 128-point constellation used with V.33 is shown in Figure 30. 209
- 215. Fundamental Concepts of Data Communications 13-5 ITU-T Modem Recommendation V.42 and V.42bis In 1988, the ITU adopted theV.42 standard error-correcting procedures for DCEs (modems). V.42 specifications address asynchronous-to-synchronous transmission conversions and error control that includes both detection and correction. V.42’s primary purpose specifies a rela- tively new modem protocol called Link Access Procedures for Modems (LAP M). LAP M is almost identical to the packet-switching protocol used with the X.25 standard. V.42bis is a specification designed to enhance the error-correcting capabilities of modems that implement the V.42 standard. Modems employing data compression schemes have proven to significantly surpass the data throughput performance of the predecessors. The V.42bis standard is capable of achieving somewhere between 3-to-1 and 4-to-1 compression ratios forASCII-coded text. The compression algorithm specified is British Telecom’s BTLZ. Throughput rates of up to 56 kbps can be achieved using V.42bis data compression. 13-6 ITU-T Modem Recommendation V.32 (V.fast) Officially adopted in 1994, V.fast is considered the next generation in data transmission. Data rates of 28.8 kbps without compression are possible using V.34. Using current data compression techniques, V.fast modems will be able to transmit data at two to three times current data rates. V.32 automatically adapts to changes in transmission-line characteristics and dynamically adjusts data rates either up or down, depending on the quality of the com- munication channel. V.34 innovations include the following: 1. Nonlinear coding, which offsets the adverse effects of system nonlinearities that produce harmonic and intermodulation distortion and amplitude proportional noise 2. Multidimensional coding and constellation shaping, which enhance data immu- nity to channel noise 3. Reduced complexity in decoders found in receivers 4. Precoding of data for more of the available bandwidth of the communications channel to be used by improving transmission of data in the outer limits of the channel where amplitude, frequency, and phase distortion are at their worst 5. Line probing, which is a technique that receive modems to rapidly determine the best correction to compensate for transmission-line impairments 13-7 ITU-T Modem Recommendation V.34 V.34 is an enhanced standard adopted by the ITU in 1996.V.34 adds 31.2 kbps and 33.6 kbps to the V.34 specification. Theoretically, V.34 adds 17% to the transmission rate; however, it is not significant enough to warrant serious consideration at this time. 13-8 ITU-T Modem Recommendation V.90 The ITU-T developed the V.90 specification in February 1998 during a meeting in Geneva, Switzerland. The V.90 recommendation is similar to 3COM’s x2 and Lucent’s K56flex in that it defines an asymmetrical data transmission technology where the upstream and down- stream data rates are not the same. V.90 allows modem downstream (receive) data rates up to 56 kbps and upstream (transmit) data rates up to 33.6 kbps. These data rates are inap- propriate in the United States and Canada, as the FCC and CRTC limit transmission rates offered by telephone companies to no more than 53 kbps. 13-9 ITU-T Modem Recommendation V.92 In2000,theITUapprovedanewmodemstandardcalledV.92.V.92offersthreeimprovements over V.90 that can be achieved only if both the transmit and receive modems and the Internet Service Provider (ISP) have V.92 compliant modems. V.92 offers (1) upstream transmission rate of 48 kbps, (2) faster call setup capabilities, and (3) incorporation of a hold option. 210
- 216. Fundamental Concepts of Data Communications QUESTIONS 1. Define data communications code. 2. Give some of the alternate names for data communications codes. 3. Briefly describe the following data communications codes: Baudot, ASCII, and EBCDIC. 4. Describe the basic concepts of bar codes. 5. Describe a discrete bar code; continuous bar code; 2D bar code. 6. Explain the encoding formats used with Code 39 and UPC bar codes. 7. Describe what is meant by error control. 8. Explain the difference between error detection and error correction. 9. Describe the difference between redundancy and redundancy checking. 10. Explain vertical redundancy checking. 11. Define odd parity; even parity; marking parity. 12. Explain the difference between no parity and ignored parity. 13. Describe how checksums are used for error detection. 14. Explain longitudinal redundancy checking. 15. Describe the difference between character and message parity. 16. Describe cyclic redundancy checking. 17. Define forward error correction. 18. Explain the difference between using ARQ and a Hamming code. 19. What is meant by character synchronization? 20. Compare and contrast asynchronous and synchronous serial data formats. 21. Describe the basic format used with asynchronous data. 22. Define the start and stop bits. 23. Describe synchronous data. 24. What is a SYN character? 25. Define and give some examples of data terminal equipment. 26. Define and give examples of data communications equipment. 27. List and describe the basic components that make up a data communications circuit. 28. Define line control unit and describe its basic functions in a data communications circuit. 29. Describe the basic functions performed by a UART. 30. Describe the operation of a UART transmitter and receiver. 31. Explain the operation of a start bit verification circuit. 32. Explain clock slippage and describe the effects of slipping over and slipping under. 33. Describe the differences between UARTs, USRTs, and USARTs. 34. List the features provided by serial interfaces. 35. Describe the purpose of a serial interface. 36. Describe the physical, electrical, and functional characteristics of the RS-232 interface. 37. Describe the RS-449 interface and give the primary differences between it and the RS-232 in- terface. 38. Describe data communications modems and explain where they are used in data communications circuits. 39. What is meant by a Bell System–compatible modem? 40. What is the difference between asynchronous and synchronous modems? 41. Define modem synchronization and list its functions. 42. Describe modem equalization. 43. Briefly describe the following ITU-T modem recommendations: V.29, V.32, V.32bis, V.32terbo, V.33, V.42, V.42bis, V.32fast, and V.34. 211
- 217. Fundamental Concepts of Data Communications PROBLEMS 1. Determine the hex codes for the following Baudot codes: C, J, 4, and /. 2. Determine the hex codes for the following ASCII codes: C, J, 4, and /. 3. Determine the hex codes for the following EBCDIC codes: C, J, 4, and /. 4. Determine the left- and right-hand UPC label format for the digit 4. 5. Determine the LRC and VRC for the following message (use even parity for LRC and odd par- ity for VCR): D A T A sp C O M M U N I C A T I O N S 6. Determine the LRC and VRC for the following message (use even parity for LRC and odd par- ity for VCR): A S C I I sp C O D E 7. Determine the BCS for the following data- and CRC-generating polynomials: G(x) x7 x4 x2 x0 1 0 0 1 0 1 0 1 P(x) x5 x4 x1 x0 1 1 0 0 1 1 8. Determine the BCC for the following data- and CRC-generating polynomials: G(x) x8 x5 x2 x0 P(x) x5 x4 x1 x0 9. How many Hamming bits are required for a single EBCDIC character? 10. Determine the Hamming bits for the ASCII character “B.” Insert the hamming bits into every other bit location starting from the left. 11. Determine the Hamming bits for the ASCII character “C” (use odd parity and two stop bits). In- sert the Hamming bits into every other location starting at the right. 12. Determine the noise margins for an RS-232 interface with driver output signal voltages of 12 V. 13. Determine the noise margins for an RS-232 interface with driver output signal voltages of 11 V. ANSWERS TO SELECTED PROBLEMS 1. C 0E, J 1A, 4 0A, / 17 3. C C3, J D1, 4 F4, / 61 5. 10100000 binary, A0 hex 7. 1000010010100000 binary, 84A0 hex 9. 4 11. Hamming bits 0010 in positions 8, 6, 4, and 2 13. 8 V 212
- 218. Data-Link Protocols and Data Communications Networks CHAPTER OUTLINE 1 Introduction 7 High-Level Data-Link Control 2 Data-Link Protocol Functions 8 Public Switched Data Networks 3 Character- and Bit-Oriented Data-Link 9 CCITT X.25 User-to-Network Interface Protocol 4 Protocols 10 Integrated Services Digital Network 4 Asynchronous Data-Link Protocols 11 Asynchronous Transfer Mode 5 Synchronous Data-Link Protocols 12 Local Area Networks 6 Synchronous Data-Link Control 13 Ethernet OBJECTIVES ■ Define data-link protocol ■ Define and describe the following data-link protocol functions: line discipline, flow control, and error control ■ Define character- and bit-oriented protocols ■ Describe asynchronous data-link protocols ■ Describe synchronous data-link protocols ■ Explain binary synchronous communications ■ Define and describe synchronous data-link control ■ Define and describe high-level data-link control ■ Describe the concept of a public data network ■ Describe the X.25 protocol ■ Define and describe the basic concepts of asynchronous transfer mode ■ Explain the basic concepts of integrated services digital network ■ Define and describe the fundamental concepts of local area networks ■ Describe the fundamental concepts of Ethernet ■ Describe the differences between the various types of Ethernet ■ Describe the Ethernet II and IEEE 802.3 frame formats From Chapter 5 of Advanced Electronic Communications Systems, Sixth Edition. Wayne Tomasi. Copyright © 2004 by Pearson Education, Inc. Published by Pearson Prentice Hall. All rights reserved. 213
- 219. 1 INTRODUCTION The primary goal of network architecture is to give users of a network the tools necessary for setting up the network and performing data flow control. A network architecture out- lines the way in which a data communications network is arranged or structured and gen- erally includes the concepts of levels or layers within the architecture. Each layer within the network consists of specific protocols or rules for communicating that perform a given set of functions. Protocols are arrangements between people or processes. A data-link protocol is a set of rules implementing and governing an orderly exchange of data between layer two de- vices, such as line control units and front-end processors. 2 DATA-LINK PROTOCOL FUNCTIONS For communications to occur over a data network, there must be at least two devices work- ing together (one transmitting and one receiving). In addition, there must be some means of controlling the exchange of data. For example, most communication between computers on networks is conducted half duplex even though the circuits that interconnect them may be capable of operating full duplex. Most data communications networks, especially local area networks, transfer data half duplex where only one device can transmit at a time. Half- duplex operation requires coordination between stations. Data-link protocols perform cer- tain network functions that ensure a coordinated transfer of data. Some data networks des- ignate one station as the control station (sometimes called the primary station). This is sometimes referred to as primary-secondary communications. In centrally controlled net- works, the primary station enacts procedures that determine which station is transmitting and which is receiving. The transmitting station is sometimes called the master station, whereas the receiving station is called the slave station. In primary-secondary networks, there can never be more than one master at a time; however, there may be any number of slave stations. In another type of network, all stations are equal, and any station can transmit at any time. This type of network is sometimes called a peer-to-peer network. In a peer-to-peer network, all sta- tions have equal access to the network, but when they have a message to transmit, they must contend with the other stations on the network for access to the transmission medium. Data-link protocol functions include line discipline, flow control, and error control. Line discipline coordinates hop-to-hop data delivery where a hop may be a computer, a net- work controller, or some type of network-connecting device, such as a router. Line disci- pline determines which device is transmitting and which is receiving at any point in time. Flow control coordinates the rate at which data are transported over a link and generally provides an acknowledgment mechanism that ensures that data are received at the desti- nation. Error control specifies means of detecting and correcting transmission errors. 2-1 Line Discipline In essence, line discipline is coordinating half-duplex transmission on a data communica- tions network. There are two fundamental ways that line discipline is accomplished in a data communications network: enquiry/acknowledgment (ENQ/ACK) and poll/select. 2-1-1 ENQ/ACK. Enquiry/acknowledgment (ENQ/ACK) is a relatively simple data-link-layer line discipline that works best in simple network environments where there is no doubt as to which station is the intended receiver. An example is a network comprised of only two stations (i.e., a two-point network) where the stations may be interconnected permanently (hardwired) or on a temporary basis through a switched network, such as the public telephone network. Before data can be transferred between stations, procedures must be invoked that es- tablish logical continuity between the source and destination stations and ensure that the Data-Link Protocols and Data Communications Networks 214
- 220. destination station is ready and capable of receiving data. These are the primary purposes of line discipline procedures. ENQ/ACK line discipline procedures determine which device on a network can initiate a transmission and whether the intended receiver is available and ready to receive a message. Assuming all stations on the network have equal access to the transmission medium, a data session can be initiated by any station using ENQ/ACK. An exception would be a receive-only device, such as most printers, which cannot initiate a ses- sion with a computer. The initiating station begins a session by transmitting a frame, block, or packet of data called an enquiry (ENQ), which identifies the receiving station. There does not seem to be any universally accepted standard definition of frames, blocks, and packets other than by size. Typically, packets are smaller than frames or blocks, although sometimes the term packet means only the information and not any overhead that may be included with the mes- sage. The terms block and frame, however, can usually be used interchangeably. In essence, the ENQ sequence solicits the receiving station to determine if it is ready to receive a message. With half-duplex operation, after the initiating station sends an ENQ, it waits for a response from the destination station indicating its readiness to receive a mes- sage. If the destination station is ready to receive, it responds with a positive acknowledg- ment (ACK), and if it is not ready to receive, it responds with a negative acknowledgment (NAK). If the destination station does not respond with an ACK or a NAK within a speci- fied period of time, the initiating station retransmits the ENQ. How many enquiries are made varies from network to network, but generally after three unsuccessful attempts to es- tablish communications, the initiating station gives up (this is sometimes called a time-out). The initiating station may attempt to establish a session later, however, after several unsuc- cessful attempts; the problem is generally referred to a higher level of authority (such as a human). A NAK transmitted by the destination station in response to an ENQ generally indi- cates a temporary unavailability, and the initiating station will simply attempt to establish a session later. An ACK from the destination station indicates that it is ready to receive data and tells the initiating station it is free to transmit its message. All transmitted message frames end with a unique terminating sequence, such as end of transmission (EOT), which indicates the end of the message frame. The destination station acknowledges all message frames received with either anACK or a NAK.AnACK transmitted in response to a received message indicates the message was received without errors, and a NAK indicates that the message was received containing errors.A NAK transmitted in response to a message is usu- ally interpreted as an automatic request for retransmission of the rejected message. Figure 1 shows how a session is established and how data are transferred using ENQ/ACK procedures. Station A initiates the session by sending an ENQ to station B. Sta- tion B responds with an ACK indicating that it is ready to receive a message. Station A transmits message frame 1, which is acknowledged by station B with an ACK. Station A then transmits message frame 2, which is rejected by station B with a NAK, indicating that the message was received with errors. Station A then retransmits message frame 2, which is received without errors and acknowledged by station B with an ACK. 2-1-2 Poll/select. The poll/select line discipline is best suited to centrally controlled data communications networks using a multipoint topology, such as a bus, where one sta- tion or device is designated as the primary or host station and all other stations are desig- nated as secondaries. Multipoint data communications networks using a single transmis- sion medium must coordinate access to the transmission medium to prevent more than one station from attempting to transmit data at the same time. In addition, all exchanges of data must occur through the primary station. Therefore, if a secondary station wishes to trans- mit data to another secondary station, it must do so through the primary station. This is anal- ogous to transferring data between memory devices in a computer using a central Data-Link Protocols and Data Communications Networks 215
- 221. Station A Station B ENQ ACK ACK NAK ACK ACK Time Time Message 3 EOT Message 2 retransmitted Message 2 Message 1 FIGURE 1 Example of ENQ/ACK line discipline processing unit (CPU) where all data are read into the CPU from the source memory and then written to the destination memory. In a poll/select environment, the primary station controls the data link, while sec- ondary stations simply respond to instructions from the primary. The primary determines which device or station has access to the transmission channel (medium) at any given time. Hence, the primary initiates all data transmissions on the network with polls and selections. A poll is a solicitation sent from the primary to a secondary to determine if the sec- ondary has data to transmit. In essence, the primary designates a secondary as a transmit- ter (i.e., the master) with a poll. A selection is how the primary designates a secondary as a destination or recipient of data. A selection is also a query from the primary to determine if the secondary is ready to receive data. With two-point networks using ENQ/ACK proce- dures, there was no need for addresses because transmissions from one station were obvi- ously intended for the other station. On multipoint networks, however, addresses are nec- essary because all transmissions from the primary go to all secondaries, and addresses are necessary to identify which secondary is being polled or selected.All secondary stations re- ceive all polls and selections transmitted from the primary. With poll/select procedures, each secondary station is assigned one or more address for identification. It is the second- aries’responsibility to examine the address and determine if the poll or selection is intended for them. The primary has no address because transmissions from all secondary stations go only to the primary. A primary can poll only one station at a time; however, it can select more than one secondary at a time using group (more than one station) or broadcast (all sta- tions) addresses. Data-Link Protocols and Data Communications Networks 216
- 222. Primary Station Station A Station B Secondary Stations Station C Poll A NAK Message Message ACK Time Selection C Selection B Poll B NAK FIGURE 2 Example of poll/select line discipline When a primary polls a secondary, it is soliciting the secondary for a message. If the secondary has a message to send, it responds to the poll with the message. This is called a positive acknowledgment to a poll. If the secondary has no message to send, it responds with a negative acknowledgment to the poll, which confirms that it received the poll but indicates that it has no messages to send at that time. This is called a negative acknowledg- ment to a poll. When a primary selects a secondary, it is identifying the secondary as a receiver. If the secondary is available and ready to receive data, it responds with an ACK. If it is not available or ready to receive data, it responds with a NAK. These are called, respectively, positive and negative acknowledgments to a selection. Figure 2 shows how polling and selections are accomplished using poll/select proce- dures. The primary polls station A, which responds with a negative acknowledgment to a poll (NAK) indicating it received the poll but has no message to send. Then the primary polls station B, which responds with a positive acknowledgment to a poll (i.e., a message). The primary then selects station B to see if it ready to receive a message. Station B responds with a positive acknowledgment to the selection (ACK), indicating that it is ready to receive a message. The primary transmits the message to station B. The primary then selects sta- tion C, which responds with a negative acknowledgment to the selection (NAK), indicating it is not ready to receive a message. Data-Link Protocols and Data Communications Networks 217
- 223. Source (transmitting) Station Destination (receiving) Station ACK Time Time Message frame 1 ACK Message frame 2 ACK Message frame 3 (EOT) Wait time Wait time Wait time FIGURE 3 Example of stop-and-wait flow control 2-2 Flow Control Flow control defines a set of procedures that tells the transmitting station how much data it can send before it must stop transmitting and wait for an acknowledgment from the desti- nation station. The amount of data transmitted must not exceed the storage capacity of the destination station’s buffer. Therefore, the destination station must have some means of in- forming the transmitting station when its buffers are nearly at capacity and telling it to tem- porarily stop sending data or to send data at a slower rate. There are two common methods of flow control: stop-and-wait and sliding window. 2-2-1 Stop-and-wait flow control. With stop-and-wait flow control, the transmit- ting station sends one message frame and then waits for an acknowledgment before send- ing the next message frame. After it receives an acknowledgment, it transmits the next frame. The transmit/acknowledgment sequence continues until the source station sends an end-of-transmission sequence. The primary advantage of stop-and-wait flow control is sim- plicity. The primary disadvantage is speed, as the time lapse between each frame is wasted time. Each frame takes essentially twice as long to transmit as necessary because both the message and the acknowledgment must traverse the entire length of the data link before the next frame can be sent. Figure 3 shows an example of stop-and-wait flow control. The source station sends message frame 1, which is acknowledged by the destination station. After stopping trans- mission and waiting for the acknowledgment, the source station transmits the next frame (message frame 2). After sending the second frame, there is another lapse in time while the destination station acknowledges reception of frame 2. The time it takes the source station to transport three frames equates to at least three times as long as it would have taken to send the message in one long frame. Data-Link Protocols and Data Communications Networks 218
- 224. 2-2-2 Sliding window flow control. With sliding window flow control, a source station can transmit several frames in succession before receiving an acknowledgment. There is only one acknowledgment for several transmitted frames, thus reducing the num- ber of acknowledgments and considerably reducing the total elapsed transmission time as compared to stop-and-wait flow control. The term sliding window refers to imaginary receptacles at the source and destination stations with the capacity of holding several frames of data. Message frames can be acknowl- edged any time before the window is filled with data. To keep track of which frames have been acknowledged and which have not, sliding window procedures require a modulo-n number- ing system where each transmitted frame is identified with a unique sequence number be- tween 0 and n 1. n is any integer value equal to 2x , where x equals the number of bits in the numbering system. With a three-bit binary numbering system, there are 23 , or eight, pos- sible numbers (0, 1, 2, 3, 4, 5, 6, and 7), and therefore the windows must have the capacity of holding n 1 (seven) frames of data. The reason for limiting the number of frames to n 1 is explained in Section 6-1-3. The primary advantage of sliding window flow control is network utilization. With fewer acknowledgments (i.e., fewer line turnarounds), considerably less network time is wasted acknowledging messages, and more time can be spent actually sending messages. The primary disadvantages of sliding window flow control are complexity and hardware capacity. Each secondary station on a network must have sufficient buffer space to hold 2(n 1) frames of data (n 1 transmit frames and n 1 receive frames), and the primary station must have sufficient buffer space to hold m(2[n 1]), where m equals the number of secondary stations on the network. In addition, each secondary must store each unacknow- ledged frame it has transmitted and keep track of the number of each unacknowledged frame it transmits and receives. The primary station must store and keep track of all unacknow- ledged frames it has transmitted and received for each secondary station on the network. 2-3 Error Control Error control includes both error detection and error correction. However, with the data-link layer, error control is concerned primarily with error detection and message retransmission, which is the most common method of error correction. With poll/select line disciplines, all polls, selections, and message transmissions end with some type of end-of-transmission sequence. In addition, all messages transported from the primary to a secondary or from a secondary to the primary are acknowledged with ACK or NAK sequences to verify the validity of the message. An ACK means the message was received with no transmission errors, and a NAK means there were errors in the received message. A NAK is an automatic call for retransmission of the last message. Error detection at the data-link layer can be accomplished with any of the methods you've learned, such as VRC, LRC, or CRC. Error correction is generally accomplished with a type of retransmission called automatic repeat request (ARQ) (sometimes called aut- omatic request for retransmission). With ARQ, when a transmission error is detected, the destination station sends a NAK back to the source station requesting retransmission of the last message frame or frames. ARQ also calls for retransmission of missing or lost frames, which are frames that either never reach the secondary or are damaged so severely that the destination station does not recognize them. ARQ also calls for retransmission of frames where the acknowledgments (either ACKs or NAKs) are lost or damaged. There are two types of ARQ: stop-and-wait and sliding window. Stop-and-wait flow control generally incorporates stop-and-wait ARQ, and sliding window flow control can implement ARQ in one of two variants: go-back-n frames or selective reject (SREJ). With go-back-n frames, the destination station tells the source station to go back n frames and re- transmit all of them, even if all the frames did not contain errors. Go-back-n requests re- transmission of the damaged frame plus any other frames that were transmitted after it. If Data-Link Protocols and Data Communications Networks 219
- 225. the second frame in a six-frame message were received in error, five frames would be re- transmitted. With selective reject, the destination station tells the source station to retrans- mit only the frame (or frames) received in error. Go-back-n is easier to implement, but it also wastes more time, as most of the frames retransmitted were not received in error. Se- lective reject is more complicated to implement but saves transmission time, as only those frames actually damaged are retransmitted. 3 CHARACTER- AND BIT-ORIENTED DATA-LINK PROTOCOLS All data-link protocols transmit control information either in separate control frames or in the form of overhead that is added to the data and included in the same frame. Data-link protocols can be generally classified as either character or bit oriented. 3-1 Character-Oriented Protocols Character-oriented protocols interpret a frame of data as a group of successive bits com- bined into predefined patterns of fixed length, usually eight bits each. Each group of bits represents a unique character. Control information is included in the frame in the form of standard characters from an existing character set, such as ASCII. Control characters con- vey important information pertaining to line discipline, flow control, and error control. With character-oriented protocols, unique data-link control characters, such as start of text (STX) and end of text (ETX), no matter where they occur in a transmission, warrant the same action or perform the same function. For example, the ASCII code 02 hex repre- sents the STX character. Start of text, no matter where 02 hex occurs within a data trans- mission, indicates that the next character is the first character of the text or information por- tion of the message. Care must be taken to ensure that the bit sequences for data-link control characters do not occur within a message unless they are intended to perform their desig- nated data-link functions. Character-oriented protocols are sometimes called byte-oriented protocols. Exam- ples of character-oriented protocols are XMODEM, YMODEM, ZMODEM, KERMIT, BLAST, IBM 83B Asynchronous Data Link Protocol, and IBM Binary Synchronous Com- munications (BSC—bisync). Bit-oriented protocols are more efficient than character- oriented protocols. 3-2 Bit-Oriented Protocols A bit-oriented protocol is a discipline for serial-by-bit information transfer over a data com- munications channel. With bit-oriented protocols, data-link control information is trans- ferred as a series of successive bits that may be interpreted individually on a bit-by-bit ba- sis or in groups of several bits rather than in a fixed-length group of n bits where n is usually the number of bits in a data character. In a bit-oriented protocol, there are no dedicated data- link control characters. With bit-oriented protocols, the control field within a frame may convey more than one control function. Bit-oriented typically convey more information into shorter frames than character- oriented protocols. The most popular bit-oriented protocol are Synchronous Data Link Communications (SDLC) and High-Level Data Link Communications (HDLC). 4 ASYNCHRONOUS DATA-LINK PROTOCOLS Asynchronous data-link protocols are relatively simple, character-oriented protocols gen- erally used on two-point networks using asynchronous data and asynchronous modems. Asynchronous protocols, such as XMODEM and YMODEM, are commonly used to facil- itate communications between two personal computers over the public switched telephone network. Data-Link Protocols and Data Communications Networks 220
- 226. *Each 8-bit character contains start and stop bits (white bars) and characters are separated from each other with gaps. S O H 8 bits 8 bits 8 bits Data field (128 bytes – fixed length) Gaps Ones Complement Sequence number Header field 8 bits C R C Error detection FIGURE 4 XMODEM frame format 4-1 XMODEM In 1979, a man named Ward Christiansen developed the first file transfer protocol designed to facilitate transferring data between two personal computers (PCs) over the public switched telephone network. Christiansen’s protocol is now called XMODEM. XMODEM is a relatively simple data-link protocol intended for low-speed applications. Although XMODEM was designed to provide communications between two PCs, it can also be used between a PC and a mainframe or host computer. XMODEM specifies a half-duplex stop-and-wait protocol using a data frame com- prised of four fields. The frame format for XMODEM contains four fields as shown in Figure 4. The four fields for XMODEM are the SOH field, header field, data field, and er- ror-detection field. The first field of an XMODEM frame is simply a one-byte start of head- ing (SOH) field. SOH is a data-link control character that is used to indicate the beginning of a header. Headers are used for conveying system information, such as the message num- ber. SOH simply indicates that the next byte is the first byte of the header. The second field is a two-byte sequence that is the actual header for the frame. The first header byte is called the sequence number, as it contains the number of the current frame being transmitted. The second header byte is simply the 2-s complement of the first byte, which is used to verify the validity of the first header byte (this is sometimes called complementary redundancy). The next field is the information field, which contains the actual user data. The information field has a maximum capacity of 128 bytes (e.g., 128 ASCII characters). The last field of the frame is an eight-bit CRC frame check sequence, which is used for error detection. Data transmission and control are quite simple with the XMODEM protocol—too simple for most modern-day data communications networks. The process of transferring data begins when the destination station sends a NAK character to the source station. Al- though NAK is the acronym for a negative acknowledgment, when transmitted by the des- tination station at the beginning of an XMODEM data transfer, it simply indicates that the destination station is ready to receive data. After the source station receives the initial NAK character, it sends the first data frame and then waits for an acknowledgment from the des- tination station. If the data are received without errors, the destination station responds with an ACK character (positive acknowledgment). If the data is received with errors, the desti- nation station responds with a NAK character, which calls for a retransmission of the data. After the originate station receives the NAK character, it retransmits the same frame. Each Data-Link Protocols and Data Communications Networks 221
- 227. time the destination station receives a frame, it responds with either a NAK or an ACK, de- pending on whether a transmission error has occurred. If the source station does not receive an ACK or NAK after a predetermined length of time, it retransmits the last frame. A time- out is treated the same as a NAK. When the destination station wishes to prematurely ter- minate a transmission, it inserts a cancel (CAN) character. 4-2 YMODEM YMODEM is a protocol similar to XMODEM except with the following exceptions: 1. The information field has a maximum capacity of 1024 bytes. 2. Two CAN characters are required to abort a transmission. 3. ITU-T-CRC 16 is used to calculate the frame check sequence. 4. Multiple frames can be sent in succession and then acknowledged with a single ACK or NAK character. SYNCHRONOUS DATA-LINK PROTOCOLS With synchronous data-link protocols, remote stations can have more than one PC or printer. A group of computers, printers, and other digital devices is sometimes called a cluster.A single line control unit (LCU) can serve a cluster with as many as 50 devices. Syn- chronous data-link protocols are generally used with synchronous data and synchronous modems and can be either character or bit oriented. One of the most common synchronous data-link protocols is IBM’s binary synchronous communications (BSC). Binary Synchronous Communications Binary synchronous communications (BSC) is a synchronous character-oriented data-link protocol developed by IBM. BSC is sometimes called bisync or bisynchronous communi- cations. With BSC, each data transmission is preceded by a unique synchronization (SYN) character as shown here: The message can be a poll, a selection, an acknowledgment, or a message containing user information. The SYN character withASCII is 16 hex and with EBCDIC 32 hex. The SYN character placesthereceiverinthecharacter(byte)modeandpreparesittoreceivedataineight-bitgroup- ings. With BSC, SYN characters are always transmitted in pairs (hence the name bisync or bi- synchronous communications). Received data are shifted serially one bit at a time through the detection circuit, where they are monitored in groups of 16 bits looking for two SYN charac- ters. Two SYN characters are used to avoid misinterpreting a random eight-bit sequence in the middle of a message with the same bit sequence as a SYN character. For example, if theASCII charactersA and b were received in succession, the following bit sequence would occur: S S Y Y message N N 0 1 0 0 0 0 0 1 A (41H) 1 hex 6 hex False STN character 0 1 1 0 0 0 1 0 b (62H) Data-Link Protocols and Data Communications Networks 5 5-1 222
- 228. P S S E P S S S S E P A Y Y O A Y Y P P N A D N N T D N N A A Q D Time fill Clearing character General poll twice ( = any device) Station polling address twice Line turnaround Leading pad Trailing pad As you can see, it appears that a SYN character has been received when in fact it has not. To avoid this situation, SYN characters are always transmitted in pairs and, consequently, if only one is detected, it is ignored. The likelihood of two false SYN characters occurring one immediately after the other is remote. 5-1-1 BSC polling sequences. BSC uses a poll/select format to control data trans- mission. There are two polling formats used with bisync: general and specific. The format for a general poll is Data-Link Protocols and Data Communications Networks where The PAD character at the beginning of the sequence is called a leading pad and is either 55 hex or AA hex (01010101 or 10101010). A leading pad is simply a string of alter- nating 1s and 0s for clock synchronization. Immediately following the leading pad are two SYN characters that establish character synchronization. The EOT character is a clearing character that places all secondary stations into the line monitor mode. The PAD character immediately following the second SYN character is simply a string of successive logic 1s that serves as a time fill, giving each of the secondary stations time to clear. The number of logic 1s transmitted during this time fill is not necessarily a mul- tiple of eight bits. Consequently, two more SYN characters are transmitted to reestab- lish character synchronization. Two station polling address (SPA) characters are trans- mitted for error detection (character redundancy). A secondary will not recognize or respond to a poll unless its SPA appears twice in succession. The two quotation marks signify that the poll is a general poll for any device at that station that has a formatted message to send. The enquiry (ENQ) character is sometimes called a format or line turn- around character because it simply completes the polling sequence and initiates a line turnaround. The PAD character at the end of the polling sequence is a trailing pad (FF). The pur- pose of the trailing pad is to ensure that the RLSD signal in the receive modem is held ac- tive long enough for the entire message to be demodulated. E N inquiry Q “ identifies a general poll S P station polling address A E O end of transmission T S Y synchronization character N P A pad D 223
- 229. P S S E P S S S S E P A Y Y O A Y Y P P D D N A D N N T D N N A A A A Q D Device address twice Table 1 Station and Device Addresses Station or Station or Device Number SPA SSA DA Device Number SPA SSA DA 0 sp sp 16 0 1 A / A 17 J 1 J 2 B S B 18 K 2 K 3 C T C 19 L 3 L 4 D U D 20 M 4 M 5 E V E 21 N 5 N 6 F W F 22 O 6 O 7 G X G 23 P 7 P 8 H Y H 24 Q 8 Q 9 I Z I 25 R 9 R 10 [ - [ 26 ] : ] 11 . , . 27 $ # $ 12 % 28 * @ * 13 ( — ( 29 ) ‘ ) 14 30 ; ; 15 ! ? ! 31 ^ ” ^ The character sequence for a specific poll is identical to a general poll except two device address (DA) characters are substituted for the two quotation marks. With a specific poll, both the station and the device address are included. Therefore, a specific poll is an invita- tion for only one specific device at a given secondary station to transmit its message.Again, two DA characters are transmitted for redundancy error detection. Table 1 lists the station polling addresses, station selection addresses, and device ad- dresses for a BSC system with a maximum of 32 stations and 32 devices. With bisync, there are only two ways in which a secondary station can respond to a poll: with a formatted message or with an ACK. The character sequence for a ACK is 5-1-2 BSC selection sequence. The format for a selection with BSC is P S S E P A Y Y O A D N N T D P S S E P S S S S E P A Y Y O A Y Y S S D D N A D N N T D N N A A A A Q D Device address Station selection address Data-Link Protocols and Data Communications Networks With BSC, there is a second polling sequence called a specific poll. The format for a specific poll is 224
- 230. heading heading Start of heading Start of text P S S S A Y Y O D N N H End of text E B P T C D X C D S T X Block of data message Block check character The sequence for a selection is very identical to a specific poll except two SSA characters are substituted for the two SPA characters. SSA stands for station selection address. All se- lections are specific, as they are for a specific device at the selected station. A secondary station can respond to a selection with either a positive or a negative ac- knowledgment. A positive acknowledgment to a selection indicates that the device selected is ready to receive. The character sequence for a positive acknowledgment is A negative acknowledgment to a selection indicates that the selected device is not ready to receive. A negative acknowledgment is called a reverse interrupt (RVI). The character se- quence for a negative acknowledgment to a selection is 5-1-3 BSC message sequence. Bisync uses stop-and-wait flow control and stop- and-wait ARQ. Formatted messages are sent from secondary stations to the primary station in response to a poll and sent from primary stations to secondary stations after the second- ary has been selected. Formatted messages use the following format: P S S D P A Y Y L 6 A D N N E D P S S D P A Y Y L 0 A D N N E D Data-Link Protocols and Data Communications Networks The block check character (BCC) uses longitudinal redundancy checking (LRC) with ASCII-coded messages and cyclic redundancy checking (CRC-16) for EBCDIC-coded messages (when CRC-16 is used, there are two BCCs). The BCC is sometimes called a block check sequence (BCS) because it does not represent a character; it is simply a se- quence of bits used for error detection. The BCC is computed beginning with the first character after SOH and continues through and includes the end of text (ETX) character. (If there is no heading, the BCC is computed beginning with the first character after start of text.) Data are transmitted in blocks or frames that are generally between 256 and 1500 bytes long. ETX is used to ter- minate the last block of a message. End of block (ETB) is used for multiple block messages to terminate all message blocks except the last one. The last block of a message is always terminated with ETX. The receiving station must acknowledge all BCCs. A positive acknowledgment to a BCC indicates that the BCC was good, and a nega- tive acknowledgment to a BCC indicates that the BCC was bad. A negative acknowledg- 225
- 231. ment is an automatic request for retransmission (ARQ). The character sequences for posi- tive and negative acknowledgments are the following: Positive responses to BCCs (messages): Negative response to BCCs (messages): where 5-1-4 BSC transparency. It is possible that a device attached to one or more of the ports of a station controller is not a computer terminal or printer. For example, a mi- croprocessor-controlled system used to monitor environmental conditions (temperature, humidity, and so on) or a security alarm system. If so, the data transferred between it and the primary are not ASCII- or EBCDIC-encoded characters. Instead, they could be mi- croprocessor op-codes or binary-encoded data. Consequently, it would be possible that an eight-bit sequence could occur within the message that is equivalent to a data-link control character. For example, if the binary code 00000011 (03 hex) occurred in a message, the controller would misinterpret it as the ASCII code for the ETX. If this happened, the con- troller would terminate the message and interpret the next sequence of bits as the BCC. To prevent this from occurring, the controller is made transparent to the data. With bi- sync, a data-link escape (DLE) character is used to achieve transparency. To place a con- troller into the transparent mode, STX is preceded by a DLE. This causes the controller to transfer the data to the selected device without searching through the message looking for data-link control characters. To come out of the transparent mode, DLE ETX is trans- mitted. To transmit a bit sequence equivalent to DLE as part of the text, it must be pre- ceded by a DLE character (i.e., DLE DLE). There are only three additional circumstances with transparent data when it is necessary to precede a character with DLE: 1. DLE ETB. Used to terminate all blocks of data except the final block. 2. DLE ITB. Used to terminate blocks of transparent text other than the final block when ITB (end of intermittent block) is used for a block-terminating character. 3. DLE SYN. With bisync, two SYN characters are inserted in the text in messages lasting longer than 1 second to ensure that the receive controller maintains char- acter synchronization. 6 SYNCHRONOUS DATA-LINK CONTROL Synchronous data-link control (SDLC) is a synchronous bit-oriented protocol developed in the 1970s by IBM for use in system network architecture (SNA) environments. SDLC was the first link-layer protocol based on synchronous, bit-oriented operation. The International N A negative acknowledgment K P S S N P A Y Y A A D N N K D P S S D P A Y Y L 0 A 1 even- numbered blocks2 D N N E D or P S S D P A Y Y L 1 A 1 odd- numbered blocks2 D N N E D Data-Link Protocols and Data Communications Networks 226
- 232. Organization for Standardization modified SDLC and created high-level data-link control (HDLC) and the International Telecommunications Union—Telecommunications Stan- dardization Sector (ITU-T) subsequently modified HDLC to create Link Access Proce- dures (LAP). The Institute of Electrical and Electronic Engineers (IEEE) modified HDLC and created IEEE 802.2. Although each of these protocol variations is important in its own domain, SDLC remains the primary SNA link-layer protocol for wide-area data networks. SDLC can transfer data simplex, half duplex, or full duplex and can support a variety of link types and topologies. SDLC can be used on point-to-point or multipoint networks over both circuit- and packet-switched networks. SDLC is a bit-oriented protocol (BOP) where there is a single control field within a message frame that performs essentially all the data-link control functions. SDLC frames are generally limited to 256 characters in length. EBCDIC was the original character language used with SDLC. There are two types of network nodes defined by SDLC: primary stations and secondary stations. There is only one primary station in an SDLC circuit, which controls data exchange on the communications channel and issues commands. All other stations on an SDLC network are secondary stations, which receive commands from the primary and return (transmit) responses to the primary station. There are three transmission states with SDLC: transient, idle, and active. The transient state exists before and after an initial transmission and after each line turnaround. A secondary station assumes the circuit is in an idle state after receiving 15 or more con- secutive logic 1s. The active state exists whenever either the primary or one of the second- ary stations is transmitting information or control signals. 6-1 SDLC Frame Format Figure 5 shows an SDLC frame format. Frames transmitted from the primary and secondary stations use exactly the same format. There are five fields included in an SDLC frame: 1. Flag field (beginning and ending) 2. Address field 3. Control field 4. Information (or text) field 5. Frame check sequence field 6-1-1 SDLC flag field. There are two flag fields per frame, each with a minimum length of one byte. The two flag fields are the beginning flag and ending flag. Flags are used for the delimiting sequence for the frame and to achieve frame and character synchroniza- tion. The delimiting sequence sets the limits of the frame (i.e., when the frame begins and when it ends). The flag is used with SDLC in a manner similar to the way SYN characters Data-Link Protocols and Data Communications Networks Flag field 8 bits Address field 8 bits Control field 8 bits Span of CRC accumulation Information field (Variable length in 8 bit groupings) Frame check sequence (CRC-16) Flag field 8 bits F A C I FCS F 01111110 7E hex 01111110 7E hex Span of zero insertion FIGURE 5 SDLC frame format 227
- 233. 01111110011111100111111001111110 0111111001111110 Address Control Text FCC Flag Flag Flag Flag Flag Flag 4. Flags are transmitted continuously during the time between frames in lieu of idle line 1s: Data-Link Protocols and Data Communications Networks 01111110 01111110 Address Beginning flag Ending flag Control Text FCC 01111110 01111110 Address Shared flag Ending flag frame N Beginning flag frame N + 1 Shared flag Ending flag frame N + 1 Beginning flag frame N + 2 Control Address Control Text FCC Text Frame N Frame N + 2 Frame N + 1 FCC 011111101111110 Beginning flag Frame N + 1 Shared 0 Shared 0 Address Address Control Text FCC Frame N+1 011111101111110 Ending flag Frame N Beginning flag Frame N + 2 Ending flag Frame N + 1 are used with bisync—to achieve character synchronization. The bit sequence for a flag is 01111110 (7E hex), which is the character “” in the EBCDIC code. There are several vari- ations to how flags are transmitted with SDLC: 1. One beginning and one ending flag for each frame: 2. The ending flag from one frame is used for the beginning flag for the next frame: 3. The last zero of an ending flag can be the first zero of the beginning flag of the next frame: 6-1-2 SDLC address field. An SDLC address field contains eight bits; therefore, 256 addresses are possible. The address 00 hex (00000000) is called the null address and is never assigned to a secondary station. The null address is used for network testing. The ad- dress FF hex (11111111) is called the broadcast address. The primary station is the only station that can transmit the broadcast address. When a frame is transmitted with the broad- cast address, it is intended for all secondary stations. The remaining 254 addresses can be used as unique station addresses intended for one secondary station only or as group ad- dresses that are intended for more than one secondary station but not all of them. In frames sent by the primary station, the address field contains the address of the sec- ondary station (i.e., the address of the destination). In frames sent from a secondary station, the address field contains the address of the secondary (i.e., the address of the station send- ing the message). The primary station has no address because all transmissions from sec- ondary stations go to the primary. 228
- 234. nr b1 P or F P or F b3 b2 b0 Function: Bit: ns b5 0 Indicates information frame b7 b6 b4 A logic 0 in the high-order bit position identifies an information frame (I-frame). With in- formation frames, the primary can select a secondary station, send formatted information, confirm previously received information frames, and poll a secondary station—with a sin- gle transmission. Bit b3 of an information frame is called a poll (P) or not-a-poll ( ) bit when sent by the primary and a final (F) or not-a-final ( ) bit when sent by a secondary. In frames sent from the primary, if the primary desires to poll the secondary (i.e., solicit it for informa- tion), the P bit in the control field is set (logic 1). If the primary does not wish to poll the secondary, the P bit is reset (logic 0). With SDLC, a secondary cannot transmit frames un- less it receives a frame addressed to it with the P bit set. This is called the synchronous re- sponse mode. When the primary is transmitting multiple frames to the same secondary, b3 is a logic 0 in all but the last frame. In the last frame, b3 is set, which demands a response from the secondary. When a secondary is transmitting multiple frames to the primary, b3 in the control field is a logic 0 in all frames except the last frame. In the last frame, b3 is set, which simply indicates that the frame is the last frame in the message sequence. In information frames, bits b4, b5, and b6 of the control field are ns bits, which are used in the transmit sequence number (ns stands for “number sent”). All information frames must be numbered.Withthreebits,thebinarynumbers000through111(0through7)canberepresented. The first frame transmitted by each station is designated frame 000, the second frame 001, and soonuptoframe111(theeighthframe),atwhichtimethecountcyclesbackto000andrepeats. SDLC uses a sliding window ARQ for error correction. In information frames, bits b0, b1, and b2 in the control field are the nr bits, which are the receive numbering sequence used to indicate the status of previously received information frames (nr stands for “number re- ceived”). The nr bits are used to confirm frames received without errors and to automatically request retransmission of information frames received with errors. The nr is the number of the next information frame the transmitting station expects to receive or the number of the next in- formation frame the receiving station will transmit. The nr confirms received frames through nr 1. Frame nr 1 is the last information frame received without a transmission error. For example, when a station transmits nr 5, it is confirming successful reception of previously unconfirmed frames up through frame 4. Together, the ns and nr bits are used for error correc- tion (ARQ). The primary station must keep track of the ns and nr for each secondary station. Each secondary station must keep track of only its ns and nr. After all frames have been con- firmed, the primary station’s ns must agree with the secondary station’s nr and vice versa. For the following example, both the primary and secondary stations begin with their nr and ns counters reset to 000. The primary begins the information exchange by sending three information frames numbered 0, 1, and 2 (i.e., the ns bits in the control character for the three frames are 000, 001, and 010). In the control character for the three frames, the F P Data-Link Protocols and Data Communications Networks 6-1-3 SDLC control field. The control field is an eight-bit field that identifies the type of frame being transmitted. The control field is used for polling, confirming previously received frames, and several other data-link management functions. There are three frame formats with SDLC: information, supervisory, and unnumbered. Information frame. With an information frame, there must be an information field, which must contain user data. Information frames are used for transmitting sequenced in- formation that must be acknowledged by the destination station. The bit pattern for the con- trol field of an information frame is 229
- 235. Primary Station Status Control Field ns nr: P/P b0 b1 b2 b3 b4 b5 b6 b7 0 0 0 0 0 0 0 0 0 0 0 00 1 0 0 0 0 0 0 0 0 1 0 02 2 0 1 3 2 0 0 0 0 1 0 1 0 0 14 ns: nr: F/F b0 b1 b2 b3 b4 b5 b6 b7 2 4 0 1 0 0 0 0 1 0 0 84 3 4 0 1 0 0 0 0 1 1 0 86 0 3 0 0 1 1 0 0 0 0 0 60 1 3 1 0 1 1 1 0 0 1 0 72 4 4 1 1 0 0 1 1 0 0 0 98 5 1 1 0 0 1 1 1 0 1 0 3A 0 1 0 0 0 1 1 0 46 4 2 1 0 1 0 1 1 0 0 0 58 4 5 0 1 0 1 0 1 0 0 0 A8 5 5 0 1 0 1 0 1 0 1 0 AA 6 5 0 1 0 1 0 1 1 0 0 AC 7 5 0 1 0 1 0 1 1 1 0 AE 0 5 1 1 0 1 1 0 0 0 0 B0 Secondary Station Status Control Field hex code hex code FIGURE 6 SDLC exchange of information frames primary transmits an nr 0 (i.e., 000). An nr 0 is transmitted because the next frame the primary expects to receive from the secondary is frame 0, which is the secondary’s present ns. The secondary responds with two information frames (ns 0 and 1). The secondary re- ceived all three frames from the primary without any errors; therefore, the nr transmitted in the secondary’s control field is 3, which is the number of the next frame the primary will send. The primary now sends information frames 3 and 4 with an nr 2, which confirms the correct reception of frames 0 and 1 from the secondary. The secondary responds with frames ns 2, 3, and 4 with an nr 4. The nr 4 confirms reception of only frame 3 from the primary (nr 1). Consequently, the primary retransmits frame 4. Frame 4 is transmit- ted together with four additional frames (ns 5, 6, 7, and 0). The primary’s nr 5, which confirms frames 2, 3, and 4 from the secondary. Finally, the secondary sends information frame 5 with an nr 1, which confirms frames 4, 5, 6, 7, and 0 from the primary. At this point, all frames transmitted have been confirmed except frame 5 from the secondary. The preceding exchange of information frames is shown in Figure 6. With SDLC, neither the primary nor the secondary station can send more than seven numbered information frames in succession without receiving a confirmation. For exam- ple, if the primary sent eight frames (ns 0, 1, 2, 3, 4, 5, 6, and 7) and the secondary re- sponded with an nr 0, it is ambiguous which frames are being confirmed. Does nr 0 Data-Link Protocols and Data Communications Networks 230
- 236. nr 5 0 P poll 1 1 1 ns 3 1 0 Information frame 0 1 0 b1 b3 b2 b0 Bit: b5 b7 b6 b4 nr 4 0 F Not a final 0 0 1 ns 7 1 0 Information frame 0 1 1 b1 b3 b2 b0 Bit: b5 b7 b6 b4 nr P or F P or F Function code Function: Indicates supervisory frame X X 1 0 b1 b3 b2 b0 Bit: b5 b7 b6 b4 mean that all eight frames were received correctly, or does it mean that frame 0 had an er- ror in it and all eight frames must be retransmitted? (All frames beginning with nr 1 must be retransmitted.) Example 1 Determine the bit pattern for the control field of an information frame sent from the primary to a sec- ondary station for the following conditions: a. Primary is sending information frame 3 (ns 3) b. Primary is polling the secondary (P 1) c. Primary is confirming correct reception of frames 2, 3, and 4 from the secondary (nr 5) Solution Data-Link Protocols and Data Communications Networks Example 2 Determine the bit pattern for the control field of an information frame sent from a secondary station to the primary for the following conditions: a. Secondary is sending information frame 7 (ns 7) b. Secondary is not sending its final frame (F 0) c. Secondary is confirming correct reception of frames 2 and 3 from the primary (nr 4) Solution Supervisory frame. With supervisory frames, an information field is not allowed. Consequently, supervisory frames cannot be used to transfer numbered information; how- ever, they can be used to assist in the transfer of information. Supervisory frames can be used to confirm previously received information frames, convey ready or busy conditions, and for a primary to poll a secondary when the primary does not have any numbered in- formation to send to the secondary. The bit pattern for the control field of a supervisory frame is A supervisory frame is identified with a 01 in bit positions b6 and b7, respectively, of the control field. With the supervisory format, bit b3 is again the poll/not-a-poll or final/not-a- final bit, and b0, b1, and b2 are the nr bits. Therefore, supervisory frames can be used by a primary to poll a secondary, and both the primary and the secondary stations can use 231
- 237. supervisory frames to confirm previously received information frames. Bits b4 and b5 in a supervisory are the function code that either indicate the receive status of the station trans- mitting the frame or request transmission or retransmission of sequenced information frames. With two bits, there are four combinations possible. The four combinations and their functions are the following: b4 b5 Receive Status 0 0 Ready to receive (RR) 0 1 Ready not to receive (RNR) 1 0 Reject (REJ) 1 1 Not used with SDLC When a primary station sends a supervisory frame with the P bit set and a status of ready to receive, it is equivalent to a general poll. Primary stations can use supervisory frames for polling and also to confirm previously received information frames without sending any information. A secondary uses the supervisory format for confirming previ- ously received information frames and for reporting its receive status to the primary. If a secondary sends a supervisory frame with RNR status, the primary cannot send it numbered information frames until that status is cleared. RNR is cleared when a secondary sends an information frame with the F bit 1 or a supervisory frame indicating RR or REJ with the F bit 0. The REJ command/response is used to confirm information frames through nr 1 and to request transmission of numbered information frames beginning with the frame number identified in the REJ frame. An information field is prohibited with a super- visory frame, and the REJ command/response is used only with full-duplex operation. Example 3 Determine the bit pattern for the control field of a supervisory frame sent from a secondary station to the primary for the following conditions: a. Secondary is ready to receive (RR) b. It is a final frame c. Secondary station is confirming correct reception of frames 3, 4, and 5 (nr 6) Solution nr 6 F final RR Supervisory frame 1 1 0 1 0 1 0 0 b1 b3 b2 b0 Bit: b5 b7 b6 b4 b1 b3 b2 b0 Bit: b5 b7 b6 b4 X P or F P or F Function: Function code Indicates unnumbered frame X X Function code X X 1 1 Data-Link Protocols and Data Communications Networks Unnumbered frame. An unnumbered frame is identified by making bits b6 and b7 in the control field both logic 1s. The bit pattern for the control field of an unnumbered frame is With an unnumbered frame, bit b3 is again either the poll/not-a-poll or final/not-a- final bit. There are five X bits (b0, b1, b2, b4, and b5) included in the control field of an un- numbered frame that contain the function code, which is used for various unnumbered commands and responses. With five bits available, there are 32 unnumbered commands/ 232
- 238. Table 2 Unnumbered Commands and Responses Binary Configuration 1 Field Resets ns b0 b7 Acronym Command Response Prohibited and nr 000 P/F 0011 UI Yes Yes No No 000 F 0111 RIM No Yes Yes No 000 P 0111 SIM Yes No Yes Yes 100 P 0011 SNRM Yes No Yes Yes 000 F 1111 DM No Yes Yes No 010 P 0011 DISC Yes No Yes No 011 F 0011 UA No Yes Yes No 100 F 0111 FRMR No Yes No No 111 F 1111 BCN No Yes Yes No 110 P/F 0111 CFGR Yes Yes No No 010 F 0011 RD No Yes Yes No 101 P/F 1111 XID Yes Yes No No 111 P/F 0011 TEST Yes Yes No No responses possible. The control field in an unnumbered frame sent from a primary station is called a command, and the control field in an unnumbered frame sent from a secondary sta- tion is called a response. With unnumbered frames, there are neither ns nor nr bits included in the control field. Therefore, numbered information frames cannot be sent or confirmed with the unnumbered format. Unnumbered frames are used to send network control and sta- tus information. Two examples of control functions are placing a secondary station on- or off-line and initializing a secondary station’s line control unit (LCU). Table 2 lists several of the more common unnumbered commands and responses. Numbered information frames are prohibited with all unnumbered frames. Therefore, user information cannot be transported with unnumbered frames and, thus, the control field in unnumbered frames does not include nr and ns bits. However, information fields contain- ing control information are allowed with the following unnumbered commands and re- sponses: UI, FRMR, CFGR, TEST, and XID. A secondary station must be in one of three modes: initialization mode, normal re- sponse mode, or normal disconnect mode. The procedures for placing a secondary station into the initialization mode are system specified and vary considerably. A secondary in the normal response mode cannot initiate unsolicited transmissions; it can transmit only in re- sponse to a frame received with the P bit set. When in the normal disconnect mode, a sec- ondary is off-line. In this mode, a secondary station will accept only the TEST, XID, CFGR, SNRM, or SIM commands from the primary station and can respond only if the P bit is set. The unnumbered commands and responses are summarized here: 1. Unnumbered information (UI). UI can be a command or a response that is used to send unnumbered information. Unnumbered information transmitted in the I-field is not acknowledged. 2. Set initialization mode (SIM). SIM is a command that places a secondary station into the initialization mode. The initialization procedure is system specified and varies from a simple self-test of the station controller to executing a complete IPL (initial program logic) program. SIM resets the ns and nr counters at the primary and secondary stations. A secondary is expected to respond to a SIM command with an unnumbered acknowledgment (UA) response. 3. Request initialization mode (RIM). RIM is a response sent by a secondary station to request the primary to send a SIM command. Data-Link Protocols and Data Communications Networks 233
- 239. 4. Set normal response mode (SNRM). SNRM is a command that places a second- ary into the normal response mode (NRM). A secondary station cannot send or receive numbered information frames unless it is in the normal response mode. Essentially, SNRM places a secondary station on-line. SNRM resets the ns and nr counters at both the primary and secondary stations. UA is the normal re- sponse to a SNRM command. Unsolicited responses are not allowed when a sec- ondary is in the NRM. A secondary station remains in the NRM until it receives a disconnect (DISC) or SIM command. 5. Disconnect mode (DM). DM is a response transmitted from a secondary station if the primary attempts to send numbered information frames to it when the sec- ondary is in the normal disconnect mode. 6. Request disconnect (RD). RD is a response sent by a secondary when it wants the primary to place it in the disconnect mode. 7. Disconnect (DISC). DISC is a command that places a secondary station in the normal disconnect mode (NDM). A secondary cannot send or receive numbered information frames when it is in the normal disconnect mode. When in the NDM, a secondary can receive only SIM or SNRM commands and can transmit only a DM response. The expected response to a DISC is UA. 8. Unnumbered acknowledgment (UA). UA is an affirmative response that indicates compliance to SIM, SNRM, or DISC commands. UA is also used to acknowl- edge unnumbered information frames. 9. Framereject(FRMR).FRMRisforreportingproceduralerrors.TheFRMRresponse is an answer transmitted by a secondary after it has received an invalid frame from the primary. A received frame may be invalid for any one of the following reasons: a. The control field contains an invalid or unassigned command. b. The amount of data in the information field exceeds the buffer space in the secondary station’s controller. c. An information field is received in a frame that does not allow information fields. d. The nr received is incongruous with the secondary’s ns, for example, if the secondary transmitted ns frames 2, 3, and 4 and then the primary responded with an nr 7. A secondary station cannot release itself from the FRMR condition, nor does it act on the frame that caused the condition. The secondary repeats the FRMR response until it receives one of the following mode-setting com- mands: SNRM, DISC, or SIM. The information field for a FRMR response must contain three bytes (24 bits) and has the following format: Data-Link Protocols and Data Communications Networks b7 b0 x x x x x x x x Control field of rejected command Secondary's present ns and nr w = 1— Invalid command x = 1— Prohibited information field received y = 1— Buffer overrun z = 1— Received nr disagrees with transmitted ns Byte 1 b7 b0 0 x x x 0 x x x ns ns Byte 2 Filler b7 b0 w x y z 0 0 0 0 Byte 3 234
- 240. 10. TEST. The TEST command/response is an exchange of frames between the pri- mary station and a secondary station. An information field may be included with the TEST command; however, it cannot be sequenced (numbered). The primary sends a TEST command to a secondary in any mode to solicit a TEST response. If an information field is included with the command, the secondary returns it with itsresponse.TheTESTcommand/responseisexchangedforlinktestingpurposes. 11. Exchange station identification (XID). XID can be a command or a response. As a command, XID solicits the identification of a secondary station. An informa- tion field can be included in the frame to convey the identification data of either the primary or the secondary. For dial-up circuits, it is often necessary that the secondary station identify itself before the primary will exchange information frames with it, although XID is not restricted only to dial-up circuits. 6-1-4 SDLC information field. All information transmitted in an SDLC frame must be in the information field (I-field), and the number of bits in the information field must be a multiple of eight.An information field is not allowed with all SDLC frames; however, the data within an information field can be user information or control information. 6-1-5 Frame Check Character (FCC) field. The FCC field contains the error de- tection mechanism for SDLC. The FCC is equivalent to the BCC used with binary syn- chronous communications (BSC). SDLC uses CRC-16 and the following generating poly- nomial: x16 + x12 + x5 + x1. Frame check characters are computed on the data in the address, control and information fields. 6-2 SDLC Loop Operation An SDLC loop operates half-duplex. The primary difference between the loop and bus con- figurations is that in a loop, all transmissions travel in the same direction on the communi- cations channel. In a loop configuration, only one station transmits at a time. The primary station transmits first, then each secondary station responds sequentially. In an SDLC loop, the transmit port of the primary station controller is connected to the receive port of the con- troller in the first down-line secondary station. Each successive secondary station is con- nected in series with the transmission path with the transmit port of the last secondary sta- tion’s controller on the loop connected to the receive port of the primary station’s controller. Figure 7 shows the physical layout for an SDLC loop. In an SDLC loop, the primary transmits sequential frames where each frame may be ad- dressed to any or all of the secondary stations. Each frame transmitted by the primary station contains an address of the secondary station to which that frame is directed. Each secondary station, in turn, decodes the address field of every frame and then serves as a repeater for all stations that are down-loop from it. When a secondary station detects a frame with its address, it copies the frame, then passes it on to the next down-loop station. All frames transmitted by the primary are returned to the primary. When the primary has completed transmitting, it fol- lows the last flag with eight consecutive logic 0s.A flag followed by eight consecutive logic 0s is called a turnaround sequence, which signals the end of the primary’s transmissions. Imme- diately following the turnaround sequence, the primary transmits continuous logic 1s, which is called the go-ahead sequence. A secondary station cannot transmit until it receives a frame ad- dress to it with the P bit set, a turnaround sequence, and then a go-ahead sequence. Once the primary has begun transmitting continuous logic 1s, it goes into the receive mode. The first down-loop secondary station that receives a frame addressed to it with the P bit set changes the go-ahead sequence to a flag, which becomes the beginning flag of that secondary sta- tion’s response frame or frames.After the secondary station has transmitted its last frame, it again becomesarepeaterfortheidleline1sfromtheprimary,whichbecomethego-aheadsequencefor thenextdown-loopsecondarystation.Thenextsecondarystationthatreceivesaframe addressed to it with the P bit set detects the turnaround sequence, any frames transmitted from up-loop Data-Link Protocols and Data Communications Networks 235
- 241. secondary stations, and then the go-ahead sequence. Each secondary station inserts its re- sponse frames immediately after the last frame transmitted by an up-loop secondary. Frames transmitted from the primary are separated from frames transmitted by the secondaries by the turnaround sequence. Without the separation, it would be impossible to tell which frames were from the primary and which were from a secondary, as their frame formats (including the address field) are identical. The cycle is completed when the primary station receives its own turnaround sequence, a series of response frames, and then the go-ahead sequence. The previously described sequence is summarized here: 1. Primary transmits sequential frames to one or more secondary stations. 2. Each transmitted frame contains a secondary station’s address. 3. After a primary has completed transmitting, it follows the last flag of the last frame with eight consecutive logic 0s (turnaround sequence) followed by continuous logic 1s (go-ahead sequence -0111111111111 - - - -). Data-Link Protocols and Data Communications Networks Copies Frame N Copies Frame N Line-control unit Station A Copies Frame A Line-control unit Station N SDLC Loop controller Primary station Tx Tx Tx Tx Cycle begins Cycle ends Go ahead Go ahead Go ahead Turnaround Turnaround Turnaround Turnaround Data frames (from primary) Line-control unit Station B Rx Rx Rx Rx GA Idle 1s Idle 1s Idle 1s Idle 1s GA A’ TA N B A A B N TA A’ B’ N’ GA A B N TA A’ B’ GA TA N B A Data frames (from primary) Data frames (from secondaries) Data frames (from primary) Data frames (from secondary A) Go ahead Data frames (from primary) Data frames (from secondaries) FIGURE 7 SDLC loop configuration 236
- 242. Data-Link Protocols and Data Communications Networks 4. The turnaround sequence alerts secondary stations of the end of the primary’s trans- missions. 5. Each secondary, in turn, decodes the address field of each frame and removes frames addressed to them. 6. Secondary stations serve as repeaters for any down-line secondary stations. 7. Secondary stations cannot transmit frames of their own unless they receive a frame with the P bit set. 8. The first secondary station that receives a frame addressed to it with the P bit set changestheseventhlogic1inthego-aheadsequencetoalogic0,thuscreatingaflag. The flag becomes the beginning flag for the secondary station’s response frames. 9. The next down-loop secondary station that receives a frame addressed to it with the P bit set detects the turnaround sequence, any frames transmitted by other up-loop secondary stations, and then the go-ahead sequence. 10. Each secondary station’s response frames are inserted immediately after the last re- peated frame. 11. The cycle is completed when the primary receives its own turnaround sequence, a series of response frames, and the go-ahead sequence. 6-2-1 SDLC loop configure command/response. The configure command/ response (CFGR) is an unnumbered command/response that is used only in SDLC loop con- figurations. CFGR contains a one-byte function descriptor (essentially a subcommand) in the information field. A CFGR command is acknowledged with a CFGR response. If the low-or- der bit of the function descriptor is set, a specified function is initiated. If it is reset, the speci- fied function is cleared. There are six subcommands that can appear in the configure command/ response function field: 1. Clear (00000000). A clear subcommand causes all previously set functions to be cleared by the secondary. The secondary’s response to a clear subcommand is an- other clear subcommand, 00000000. 2. Beacon test (BCN) (0000000X). The beacon test subcommand causes the sec- ondary receiving it to turn on (00000001) or turn off (00000000) its carrier. The beacon response is called a carrier, although it is not a carrier in the true sense of the word. The beacon test command causes a secondary station to begin trans- mitting a beacon response, which is not a carrier. However, if modems were used in the circuit, the beacon response would cause the modem’s carrier to turn on. The beacon test is used to isolate open-loop continuity problems. In addition, whenever a secondary station detects a loss of signal (either data or idle line ones), it automatically begins to transmit its beacon response. The secondary will continue transmitting the beacon until the loop resumes normal status. 3. Monitor mode (0000010X). The monitor command (00000101) causes the ad- dressed secondary station to place itself into the monitor (receive-only) mode. Once in the monitor mode, a secondary cannot transmit until it receives either a monitor mode clear (00000100) or a clear (00000000) subcommand. 4. Wrap (0000100X). The wrap command (00001001) causes a secondary station to loop its transmissions directly to its receiver input. The wrap command places the secondary effectively off-line for the duration of the test. A secondary sta- tion takes itself out of the wrap mode when it receives a wrap clear (00001000) or clear (00000000) subcommand. 5. Self-test (0000101X). The self-test subcommand (00001011) causes the ad- dressed secondary to initiate a series of internal diagnostic tests. When the tests are completed, the secondary will respond. If the P bit in the configure command is set, the secondary will respond following completion of the self-test or at its earliest opportunity. If the P bit is reset, the secondary will respond following completion of the test to the next poll-type frame it receives from the primary. 237
- 243. After zero deletion at the receive end: 01111110 01101111 11010011 1110001100110101 01111110 Beginning Address Control Frame check Ending flag character flag 6-3 Message Abort Message abort is used to prematurely terminate an SDLC frame. Generally, this is done only to accommodatehigh-prioritymessages,suchasemergencylinkrecoveryprocedures.Amessage abort is any occurrence of seven to 14 consecutive logic 1s. Zeros are not inserted in an abort sequence. A message abort terminates an existing frame and immediately begins the higher- Data-Link Protocols and Data Communications Networks 01111110 01101111 101010011 11100001100110101 01111110 Beginning flag Ending flag Address Control Inserted zeros Frame check character All other transmissions are ignored by the secondary while it is performing a self-test; however, the secondary will repeat all frames received to the next down-loop station. The secondary reports the results of the self-test by setting or clearing the low-order bit (X) of its self-test response. A logic 1 means that the tests were unsuccessful, and a logic 0 means that they were successful. 6. Modified link test (0000110X). If the modified link test function is set (X bit set), the secondary station will respond to a TEST command with a TEST response that has an information field containing the first byte of the TEST command in- formation field repeated n times. The number n is system specified. If the X bit is reset, the secondary station will respond with a zero-length information field. The modified link test is an optional subcommand and is used only to provide an alternative form of link test to that previously described for the TEST command. 6-2-2 SDLC transparency. With SDLC, the flag bit sequence (01111110) can occur within a frame where it is not intended to be a flag. For instance, within the address, control, or information fields, a combination of one or more bits from one character com- bined with one or more bits from an adjacent character could produce a 01111110 pat- tern. If this were to happen, the receive controller would misinterpret the sequence for a flag, thus destroying the frame. Therefore, the pattern 01111110 must be prohibited from occurring except when it is intended to be a flag. One solution to the problem would be to prohibit certain sequences of characters from occurring, which would be difficult to do. A more practical solution would be to make a receiver transparent to all data located between beginning and ending flags. This is called transparency. The transparency mechanism used with SDLC is called zero-bit insertion or zero stuffing. With zero-bit insertion, a logic 0 is automatically inserted after any occurrence of five consecutive logic 1s except in a designated flag sequence (i.e., flags are not zero inserted). When five consecutive logic 1s are received and the next bit is a 0, the 0 is automatically deleted or removed. If the next bit is a logic 1, it must be a valid flag. An example of zero insertion/deletion is shown here: Original frame bits at the transmit station: 01111110 01101111 11010011 1110001100110101 01111110 Beginning Address Control Frame check Ending flag character flag After zero insertion but prior to transmission: 238
- 244. Original data NRZI encoded data *I indicates to invert 1 1 0 0 0 0 0 1 0 0 0 1 1 1 0 I I I I I I I I I 1 1 FIGURE 8 NRZI encoding priority frame. If more than 14 consecutive logic 1s occur in succession, it is considered an idle line condition. Therefore, 15 or more contiguous logic 1s places the circuit into the idle state. 6-4 Invert-on-Zero Encoding With binary synchronous transmission such as SDLC, transmission and reception of data must be time synchronized to enable identification of sequential binary digits. Synchronous data communications assumes that bit or time synchronization is provided by either the DCE or the DTE. The master transmit clock can come from the DTE or, more likely, the DCE. However, the receive clock must be recovered from the data by the DCE and then transferred to the DTE. With synchronous data transmission, the DTE receiver must sample the incom- ing data at the same rate that it was outputted from the transmit DTE. Although minor vari- ations in timing can exist, the receiver in a synchronous modem provides data clock recov- ery and dynamically adjusted sample timing to keep sample times midway between bits. For a DCE to recover the data clock, it is necessary that transitions occur in the data. Traditional unipolar (UP) logic levels, such as TTL (0 V and 5 V), do not provide transitions for long strings of logic 0s or logic 1s. Therefore, they are inadequate for clock recovery without placing restrictions on the data. Invert-on-zero coding is the encoding scheme used with SDLC because it guarantees at least one transition in the data for every seven bits trans- mitted. Invert-on-zero coding is also called NRZI (nonreturn-to-zero inverted). With NRZI encoding, the data are encoded in the controller at the transmit end and then decoded in the controller at the receive end. Figure 8 shows examples of NRZI encoding and decoding. The encoded waveform is unchanged by 1s in the NRZI encoder. However, logic 0s cause the encoded transmission level to invert from its previous state (i.e., either from a high to a low or from a low to a high). Consequently, consecutive logic 0s are converted to an al- ternating high/low sequence. With SDLC, there can never be more than six logic 1s in suc- cession (a flag). Therefore, a high-to-low transition is guaranteed to occur at least once every seven bits transmitted except during a message abort or an idle line condition. In a NRZI de- coder, whenever a high/low or low/high transition occurs in the received data, a logic 0 is gen- erated. The absence of a transition simply generates a logic 1. In Figure 8, a high level is as- sumed prior to encoding the incoming data. NRZI encoding was originally intended for asynchronous modems that do not have clock recovery capabilities. Consequently, the receive DTE must provide time synchroniza- tion, which is aided by using NRZI-encoded data. Synchronous modems have built-in scram- bler and descrambler circuits that ensure transitions in the data and, thus, NRZI encoding is unnecessary. The NRZI encoder/decoder is placed in between the DTE and the DCE. 7 HIGH-LEVEL DATA-LINK CONTROL In 1975, the International Organization for Standardization (ISO) defined several sets of substandards that, when combined, are called high-level data-link control (HDLC). HDLC is a superset of SDLC; therefore, only the added capabilities are explained. Data-Link Protocols and Data Communications Networks 239
- 245. Data-Link Protocols and Data Communications Networks HDLC comprises three standards (subdivisions) that outline the frame structure, con- trol standards, and class of operation for a bit-oriented data-link control (DLC): 1. ISO 3309 2. ISO 4335 3. ISO 7809 7-1 ISO 3309 The ISO 3309 standard defines the frame structure, delimiting sequence, transparency mechanism, and error-detection method used with HDLC. With HDLC, the frame structure and delimiting sequence are essentially the same as with SDLC. An HDLC frame includes a beginning flag field, an address field, a control field, an information field, a frame check character field, and an ending flag field. The delimiting sequence with HDLC is a binary 01111110, which is the same flag sequence used with SDLC. However, HDLC computes the frame check characters in a slightly different manner. HDLC can use either CRC-16 with a generating polynomial specified by CCITT V.41 for error detection. At the transmit station, the CRC characters are computed such that when included in the FCC computa- tions at the receive end, the remainder for an errorless transmission is always FOB8. HDLC also offers an optional 32-bit CRC checksum. HDLC has extended addressing capabilities. HDLC can use an eight-bit address field or an extended addressing format, which is virtually limitless. With extended addressing, the ad- dress field may be extended recursively. If b0 in the address field is a logic 1, the seven remain- ing bits are the secondary’s address (the ISO defines the low-order bit as b0,whereas SDLC des- ignates the high-order bit as b0). If b0 is a logic 0, the next byte is also part of the address. If b0 of the second byte is a logic 0, a third address byte follows and so on until an address byte with a logic 1 for the low-order bit is encountered. Essentially, there are seven bits available in each ad- dress byte for address encoding. An example of a three-byte extended addressing scheme is shown in the following. Bit b0 in the first two bytes of the address field are logic 0s, indicating that one or more additional address bytes follow. However, b0 in the third address byte is a logic 1, which terminates the address field. There are a total of 21 address bits (seven in each byte): 7-2 Information Field HDLC permits any number of bits in the information field of an information command or response. With HDLC, any number of bits may be used for a character in the I field as long as all characters have the same number of bits. 7-3 Elements of Procedure The ISO 3309 standard defines the elements of procedure for HDLC. The control and in- formation fields have increased capabilities over SDLC, and there are two additional oper- ational modes allowed with HDLC. 7-3-1 Control field. With HDLC, the control field can be extended to 16 bits. Seven bits are for the ns, and seven bits are for the nr. Therefore, with the extended control format, there can be a maximum of 127 outstanding (unconfirmed) frames at any given time. In essence, a primary station can send 126 successive information frames to a secondary sta- tion with the P bit 0 before it would have to send a frame with the P bit 1. With HDLC, the supervisory format includes a fourth status condition: selective re- ject (SREJ). SREJ is identified by two logic 1s in bit positions b4 and b5 of a supervisory 01111110 0XXXXXXX 0XXXXXXX 1XXXXXXX Flag 1st address byte 2nd address byte Three-byte address field 3rd address byte b0 = 0 b0 = 0 b0 = 1 240
- 246. control field. With SREJ, a single frame can be rejected. A SREJ calls for the retransmis- sion of only one frame identified by the three-bit nr code. A REJ calls for the retransmis- sion of all frames beginning with frame identified by the three-bit nr code. For example, the primary sends I frames ns 2, 3, 4, and 5. If frame 3 were received in error, a REJ with an nr of 3 would call for a retransmission of frames 3, 4, and 5. However, a SREJ with an nr of 3 would call for the retransmission of only frame 3. SREJ can be used to call for the re- transmission of any number of frames except only one at a time. 7-3-2 HDLC operational modes. SDLC specifies only one operational mode, called the normal response mode (NRM), which allows secondaries to communicate with the primary only after the primary has given the secondary permission to transmit. With SDLC, when a station is logically disconnected from the network, it is said to be in the normal disconnect mode. HDLC has two additional operational modes: asynchronous response mode (ARM) and asynchronous balanced mode (ABM). With ARM, secondary stations are allowed to send unsolicited responses (i.e., communicate with the primary without permission). To transmit, a secondary does not need to have received a frame from the primary with the P bit set. However, if a secondary receives a frame with the P bit set, it must respond with a frame with the F bit set. HDLC also specifies an asynchronous disconnect mode, which is identical to the normal disconnect mode except that the secondary can initiate an asyn- chronous DM or RIM response at any time. The ISO 7809 standard combines previous standards 6159 (unbalanced) and 6256 (balanced) and outlines the class of operation necessary to establish the link-level protocol. Unbalanced operation is a class of operation logically equivalent to a multipoint private- line circuit with a polling environment. There is a single primary station responsible for central control of the network. Data transmission may be either half or full duplex. Asynchronous balanced mode is a mode of operation logically equivalent to a two- point private-line circuit where each station has equal data-link responsibilities (a station can operate as a primary or as a secondary), which enables a station to initiate data transmission without receiving permission from any other station. Channel access is accomplished through contention on a two-wire circuit using the asynchronous response mode. Data trans- mission is half duplex on a two-wire circuit or full duplex over a four-wire circuit. 8 PUBLIC SWITCHED DATA NETWORKS A public switched data network (PDN or PSDN) is a switched data communications network similar to the public telephone network except a PDN is designed for transferring data only. A public switched data network is comprised of one or more wide-area data networks designed to provide access to a large number of subscribers with a wide variety of computer equipment. The basic principle behind a PDN is to transport data from a source to a destination through a network of intermediate switching nodes and transmission media. The switching nodes are not concerned with the content of the data, as their purpose is to provide end sta- tions access to transmission media and other switching nodes that will transport data from node to node until it reaches its final destination. Figure 9 shows a public switched data network comprised of several switching nodes interconnected with transmission links (channels). The end-station devices can be personal computers, servers, mainframe com- puters, or any other piece of computer hardware capable of sending or receiving data. End stations are connected to the network through switching nodes. Data enter the network where they are routed through one or more intermediate switching nodes until reaching their destination. Some switching nodes connect only to other switching nodes (sometimes called tandem switching nodes or switchers switches), while some switching nodes are connected to end stations as well. Node-to-node communications links generally carry multiplexed data Data-Link Protocols and Data Communications Networks 241
- 247. Server Personal computer Personal computer Personal computer Personal computer Public data network Personal computer Mainframe computer Gateway to other public data networks Switching node Switching node Switching node Switching node Switching node Switching node Switching node Switching node FIGURE 9 Public switched data network (usually time-division multiplexing). Public data networks are not direct connected; that is, they do not provide direct communications links between every possible pair of nodes. Public switched data networks combine the concepts of value-added networks (VANs) and packet switching networks. 8-1 Value-Added Network A value-added network “adds value” to the services or facilities provided by a common car- rier to provide new types of communication services. Examples of added values are error control, enhanced connection reliability, dynamic routing, failure protection, logical multi- plexing, and data format conversions. A VAN comprises an organization that leases com- munications lines from common carriers such as ATT and MCI and adds new types of communications services to those lines. Examples of value-added networks are GTE Tel- net, DATAPAC, TRANSPAC, and Tymnet Inc. 8-2 Packet Switching Network Packet switching involves dividing data messages into small bundles of information and transmitting them through communications networks to their intended destinations using computer-controlled switches. Three common switching techniques are used with public data networks: circuit switching, message switching, and packet switching. 8-2-1 Circuit switching. Circuit switching is used for making a standard telephone call on the public telephone network. The call is established, information is transferred, and then the call is disconnected. The time required to establish the call is called the setup time. Once the call has been established, the circuits interconnected by the network Data-Link Protocols and Data Communications Networks 242
- 248. switches are allocated to a single user for the duration of the call. After a call has been es- tablished, information is transferred in real time. When a call is terminated, the circuits and switches are once again available for another user. Because there are a limited number of circuits and switching paths available, blocking can occur. Blocking is the inability to com- plete a call because there are no facilities or switching paths available between the source and destination locations. When circuit switching is used for data transfer, the terminal equipment at the source and destination must be compatible; they must use compatible modems and the same bit rate, character set, and protocol. A circuit switch is a transparent switch. The switch is transparent to the data; it does nothing more than interconnect the source and destination terminal equipment. A circuit switch adds no value to the circuit. 8-2-2 Message switching. Message switching is a form of store-and-forward net- work. Data, including source and destination identification codes, are transmitted into the network and stored in a switch. Each switch within the network has message storage ca- pabilities. The network transfers the data from switch to switch when it is convenient to do so. Consequently, data are not transferred in real time; there can be a delay at each switch. With message switching, blocking cannot occur. However, the delay time from message transmission to reception varies from call to call and can be quite long (possibly as long as 24 hours). With message switching, once the information has entered the net- work, it is converted to a more suitable format for transmission through the network.At the receive end, the data are converted to a format compatible with the receiving data terminal equipment. Therefore, with message switching, the source and destination data terminal equipment do not need to be compatible. Message switching is more efficient than circuit switching because data that enter the network during busy times can be held and transmit- ted later when the load has decreased. A message switch is a transactional switch because it does more than simply trans- fer the data from the source to the destination. A message switch can store data or change its format and bit rate, then convert the data back to their original form or an entirely dif- ferent form at the receive end. Message switching multiplexes data from different sources onto a common facility. 8-2-3 Packet switching. With packet switching, data are divided into smaller seg- ments, called packets, prior to transmission through the network. Because a packet can be held in memory at a switch for a short period of time, packet switching is sometimes called a hold-and-forward network. With packet switching, a message is divided into packets, and each packet can take a different path through the network. Consequently, all packets do not necessarily arrive at the receive end at the same time or in the same order in which they were transmitted. Because packets are small, the hold time is generally quite short, message transfer is near real time, and blocking cannot occur. However, packet switching networks require complex and expensive switching arrangements and complicated protocols. A packet switch is also a transactional switch. Circuit, message, and packet switching tech- niques are summarized in Table 3. 9 CCITT X.25 USER-TO-NETWORK INTERFACE PROTOCOL In 1976, the CCITT designated the X.25 user interface as the international standard for packet network access. Keep in mind that X.25 addresses only the physical, data-link, and network layers in the ISO seven-layer model. X.25 uses existing standards when possible. For example, X.25 specifies X.21, X.26, and X.27 standards as the physical interface, which correspond to EIA RS-232, RS-423A, and RS-422A standards, respectively. X.25 defines HDLC as the international standard for the data-link layer and the American Na- tional Standards Institute (ANSI) 3.66 Advanced Data Communications Control Procedures Data-Link Protocols and Data Communications Networks 243
- 249. Table 4 LAPB Commands Bit Number Command 8 7 6 5 4 3 2 1 1 (information) nr P ns 0 RR (receiver ready) nr P 0 0 0 1 RNR (receiver not ready) nr P 0 1 0 1 REJ (reject) nr P 1 0 0 1 SABM (set asynchronous balanced mode) 0 0 1 P 1 1 1 1 DISC (disconnect) 0 1 0 P 0 0 1 1 (ADCCP) as the U.S. standard. ANSI 3.66 and ISO HDLC were designed for private-line data circuits with a polling environment. Consequently, the addressing and control proce- dures outlined by them are not appropriate for packet data networks. ANSI 3.66 and HDLC were selected for the data-link layer because of their frame format, delimiting sequence, transparency mechanism, and error-detection method. At the link level, the protocol specified by X.25 is a subset of HDLC, referred to as Link Access Procedure Balanced (LAPB). LAPB provides for two-way, full-duplex com- munications between DTE and DCE at the packet network gateway. Only the address of the DTE or DCE may appear in the address field of a LAPB frame. The address field refers to a link address, not a network address. The network address of the destination terminal is embedded in the packet header, which is part of the information field. Tables 4 and 5 show the commands and responses, respectively, for an LAPB frame. During LAPB operation, most frames are commands.A response frame is compelled only when a command frame is received containing a poll (P bit) 1. SABM/UA is a command/ response pair used to initialize all counters and timers at the beginning of a session. Simi- larly, DISC/DM is a command/response pair used at the end of a session. FRMR is a re- sponse to any illegal command for which there is no indication of transmission errors ac- cording to the frame check sequence field. Information (I) commands are used to transmit packets. Packets are never sent as re- sponses. Packets are acknowledged using ns and nr just as they were in SDLC. RR is sent by a station when it needs to respond (acknowledge) something but has no information packets to send. A response to an information command could be RR with F 1. This pro- cedure is called checkpointing. Data-Link Protocols and Data Communications Networks Table 3 Switching Summary Circuit Switching Message Switching Packet Switching Dedicated transmission path No dedicated transmission path No dedicated transmission path Continuous transmission of data Transmission of messages Transmission of packets Operates in real time Not real time Near real time Messages not stored Messages stored Messages held for short time Path established for entire Route established for each Route established for each message message packet Call setup delay Message transmission delay Packet transmission delay Busy signal if called party busy No busy signal No busy signal Blocking may occur Blocking cannot occur Blocking cannot occur User responsible for Network responsible for lost Network may be responsible message-loss protection messages for each packet but not for entire message No speed or code conversion Speed and code conversion Speed and code conversion Fixed bandwidth transmission Dynamic use of bandwidth Dynamic use of bandwidth (i.e., fixed information capacity) No overhead bits after initial Overhead bits in each message Overhead bits in each packet setup delay 244
- 250. Table 5 LAPB Responses Bit Number Response 8 7 6 5 4 3 2 1 RR (receiver ready) nr F 0 0 0 1 RNR (receiver not ready) nr F 0 1 0 1 REJ (reject) nr F 1 0 0 1 UA (unnumbered acknowledgment) 0 1 1 F 0 0 1 1 DM (disconnect mode) 0 0 0 F 1 1 1 1 FRMR (frame rejected) 1 0 0 F 0 1 1 1 REJ is another way of requesting transmission of frames. RNR is used for the flow con- trol to indicate a busy condition and prevents further transmissions until cleared with an RR. The network layer of X.25 specifies three switching services offered in a switched data network: permanent virtual circuit, virtual call, and datagram. 9-1 Permanent Virtual Circuit Apermanent virtual circuit (PVC) is logically equivalent to a two-point dedicated private-line circuit except slower.A PVC is slower because a hardwired, end-to-end connection is not pro- vided. The first time a connection is requested, the appropriate switches and circuits must be established through the network to provide the interconnection. A PVC identifies the routing betweentwopredeterminedsubscribersofthenetworkthatisusedforallsubsequentmessages. With a PVC, a source and destination address are unnecessary because the two users are fixed. 9-2 Virtual Call A virtual call (VC) is logically equivalent to making a telephone call through the DDD net- work except no direct end-to-end connection is made. A VC is a one-to-many arrangement. Any VC subscriber can access any other VC subscriber through a network of switches and communication channels. Virtual calls are temporary virtual connections that use common usage equipment and circuits. The source must provide its address and the address of the destination before a VC can be completed. 9-3 Datagram A datagram (DG) is, at best, vaguely defined by X.25 and, until it is completely outlined, has very limited usefulness. With a DG, users send small packets of data into the network. Each packet is self-contained and travels through the network independent of other packets of the same message by whatever means available. The network does not acknowledge packets, nor does it guarantee successful transmission. However, if a message will fit into a single packet, a DG is somewhat reliable. This is called a single-packet-per-segment protocol. 9-4 X.25 Packet Format A virtual call is the most efficient service offered for a packet network. There are two packet formats used with virtual calls: a call request packet and a data transfer packet. 9-4-1 Call request packet. Figure 10 shows the field format for a call request packet. The delimiting sequence is 01111110 (an HDLC flag), and the error-detection/ correction mechanism is CRC-16 with ARQ. The link address field and the control field have little use and, therefore, are seldom used with packet networks. The rest of the fields are defined in sequence. Format identifier. The format identifier identifies whether the packet is a new call re- quest or a previously established call. The format identifier also identifies the packet numbering sequence (either 0–7 or 0–127). Logical channel identifier (LCI). The LCI is a 12-bit binary number that identifies the source and destination users for a given virtual call. After a source user has Data-Link Protocols and Data Communications Networks 245
- 251. gained access to the network and has identified the destination user, they are as- signed an LCI. In subsequent packets, the source and destination addresses are un- necessary; only the LCI is needed. When two users disconnect, the LCI is relin- quished and can be reassigned to new users. There are 4096 LCIs available. Therefore, there may be as many as 4096 virtual calls established at any given time. Packet type. This field is used to identify the function and the content of the packet (new request, call clear, call reset, and so on). Calling address length. This four-bit field gives the number of digits (in binary) that appear in the calling address field. With four bits, up to 15 digits can be specified. Called address length. This field is the same as the calling address field except that it identifies the number of digits that appear in the called address field. Called address. This field contains the destination address. Up to 15 BCD digits (60 bits) can be assigned to a destination user. Calling address. This field is the same as the called address field except that it con- tains up to 15 BCD digits that can be assigned to a source user. Facilities length field. This field identifies (in binary) the number of eight-bit octets present in the facilities field. Facilities field. This field contains up to 512 bits of optional network facility infor- mation, such as reverse billing information, closed user groups, and whether it is a simplex transmit or simplex receive connection. Protocol identifier. This 32-bit field is reserved for the subscriber to insert user-level protocol functions such as log-on procedures and user identification practices. User data field. Up to 96 bits of user data can be transmitted with a call request packet. These are unnumbered data that are not confirmed. This field is generally used for user passwords. 9-4-2 Data transfer packet. Figure 11 shows the field format for a data transfer packet. A data transfer packet is similar to a call request packet except that a data transfer packet has considerably less overhead and can accommodate a much larger user data field. The data transfer packet contains a send-and-receive packet sequence field that was not in- cluded with the call request format. The flag, link address, link control, format identifier, LCI, and FCS fields are identi- cal to those used with the call request packet. The send and receive packet sequence fields are described as follows: Send packet sequence field. The P(s) field is used in the same manner that the ns and nr sequences are used with SDLC and HDLC. P(s) is analogous to ns, and P(r) is analogous Data-Link Protocols and Data Communications Networks Flag 8 bits Link address field 8 bits Link control field 8 bits Format identifier 4 bits Logical channel identifier 12 bits Calling address up to 60 bits Called address up to 60 bits Called address length 4 bits Calling address length 4 bits Packet type 8 bits Null field 2 zeros Facilities length field 6 bits Facilities field up to 512 bits Protocol ID 32 bits Flag 8 bits Frame check sequence 16 bits User data up to 96 bits FIGURE 10 X.25 call request packet format 246
- 252. Flag 8 bits Link address field 8 bits Link control field 8 bits Format identifier 4 bits Logical channel identifier 12 bits Null field 5 or 1 zeros Send packet sequence number P(s) 3 or 7 bits Receive packet sequence number P(r) 3 or 7 bits Null field 5 or 1 zeros User data up to 1024 bits Flag 8 bits Frame check sequence 16 bits FIGURE 11 X.25 data transfer packet format to nr. Each successive data transfer packet is assigned the next P(s) number in sequence. The P(s) can be a 14- or seven-bit binary number and, thus, number packets from either 0–7 or 0–127. The numbering sequence is identified in the format identifier. The send packet field always contains eight bits, and the unused bits are reset. Receive packet sequence field. P(r) is used to confirm received packets and call for retransmission of packets received in error (ARQ). The I field in a data transfer packet can have considerably more source information than an I field in a call re- quest packet. 9-5 The X Series of Recommended Standards X.25 is part of the X series of ITU-T-recommended standards for public data networks. The X series is classified into two categories: X.1 through X.39, which deal with services and facilities, terminals, and interfaces, and X.40 through X.199, which deal with network ar- chitecture, transmission, signaling, switching, maintenance, and administrative arrange- ments. Table 6 lists the most important X standards with their titles and descriptions. 10 INTEGRATED SERVICES DIGITAL NETWORK The data and telephone communications industry is continually changing to meet the de- mands of contemporary telephone, video, and computer communications systems. Today, more and more people have a need to communicate with each other than ever before. In or- der to meet these needs, old standards are being updated and new standards developed and implemented almost on a daily basis. The Integrated Services Digital Network (ISDN) is a proposed network designed by the major telephone companies in conjunction with the ITU-T with the intent of pro- viding worldwide telecommunications support of voice, data, video, and facsimile in- formation within the same network (in essence, ISDN is the integrating of a wide range of services into a single multipurpose network). ISDN is a network that proposes to in- terconnect an unlimited number of independent users through a common communica- tions network. To date, only a small number of ISDN facilities have been developed; however, the telephone industry is presently implementing an ISDN system so that in the near future, subscribers will access the ISDN system using existing public telephone and data networks. The basic principles and evolution of ISDN have been outlined by the International Telecommunication Union-Telephony (ITU-T) in its recommendation ITU-T I.120 (1984). ITU-T I.120 lists the following principles and evolution of ISDN. Data-Link Protocols and Data Communications Networks 247
- 253. 10-1 Principles of ISDN The main feature of the ISDN concept is to support a wide range of voice (telephone) and nonvoice (digital data) applications in the same network using a limited number of stan- dardized facilities. ISDNs support a wide variety of applications, including both switched and nonswitched (dedicated) connections. Switched connections include both circuit- and packet-switched connections and their concatenations. Whenever practical, new services introduced into an ISDN should be compatible with 64-kbps switched digital connections. The 64-kbps digital connection is the basic building block of ISDN. An ISDN will contain intelligence for the purpose of providing service features, maintenance, and network management functions. In other words, ISDN is expected to pro- vide services beyond the simple setting up of switched circuit calls. A layered protocol structure should be used to specify the access procedures to an ISDN and can be mapped into the open system interconnection (OSI) model. Standards al- ready developed for OSI-related applications can be used for ISDN, such as X.25 level 3 for access to packet switching services. It is recognized that ISDNs may be implemented in a variety of configurations ac- cording to specific national situations. This accommodates both single-source or competi- tive national policy. 10-2 Evolution of ISDN ISDNs will be based on the concepts developed for telephone ISDNs and may evolve by progressively incorporating additional functions and network features including those of any other dedicated networks such as circuit and packet switching for data so as to provide for existing and new services. The transition from an existing network to a comprehensive ISDN may require a period of time extending over one or more decades. During this period, arrangements must be devel- oped for the internetworking of services on ISDNs and services on other networks. Data-Link Protocols and Data Communications Networks Table 6 ITU-T X Series Standards X.1 International user classes of service in public data networks. Assigns numerical class desig- nations to different terminal speeds and types. X.2 International user services and facilities in public data networks. Specifies essential and addi- tional services and facilities. X.3 Packet assembly/disassembly facility (PAD) in a public data network. Describes the packet assembler/disassembler, which normally is used at a network gateway to allow connection of a start/stop terminal to a packet network. X.20bis Use on public data networks of DTE designed for interfacing to asynchronous full-duplex V-series modems. Allows use of V.24/V.28 (essentially the same as EIA RS-232). X.21bis Use on public data networks of DTE designed for interfacing to synchronous full-duplex V-series modems. Allows use of V.24/V.28 (essentially the same as EIA RS-232) or V.35. X.25 Interface between DTE and DCE for terminals operating in the packet mode on public data networks. Defines the architecture of three levels of protocols existing in the serial inter- face cable between a packet mode terminal and a gateway to a packet network. X.28 DTE/DCE interface for a start/stop mode DTE accessing the PAD in a public data network situated in the same country. Defines the architecture of protocols existing in a serial inter- face cable between a start/stop terminal and an X.3 PAD. X.29 Procedures for the exchange of control information and user data between a PAD and a packet mode DTE or another PAD. Defines the architecture of protocols behind the X.3 PAD, either between two PADs or between a PAD and a packet mode terminal on the other side of the network. X.75 Terminal and transit call control procedures and data transfer system on international circuits between packet-switched data networks. Defines the architecture of protocols between two public packet networks. X.121 International numbering plan for public data networks. Defines a numbering plan including code assignments for each nation. 248
- 254. Private branch exchange (PBX) Packet data network Circuit switching network Other networks Customer ISDN interface Databases ATM network Other services Alarm system ISDN central office Digital pipes to other customers Digital pipes to other networks and services Telephone Local area network (LAN) Customer's premise Personal computer Digital pipe (Local subscriber loop with ISDN channel structure) Digital pipes to other customers FIGURE 12 Subscriber’s conceptual view of ISDN In the evolution toward an ISDN, digital end-to-end connectivity will be obtained via plant and equipment used in existing networks, such as digital transmission, time-division multiplex, and/or space-division multiplex switching. Existing relevant recommendations for these constituent elements of an ISDN are contained in the appropriate series of rec- ommendations of ITU-T and CCIR. In the early stages of the evolution of ISDNs, some interim user-network arrange- ments may need to be adopted in certain countries to facilitate early penetration of digital service capabilities. An evolving ISDN may also include at later stages switched connec- tions at bit rates higher and lower than 64 kbps. 10-3 Conceptual View of ISDN Figure 12 shows a view of how ISDN can be conceptually viewed by a subscriber (customer) of the system. Customers gain access to the ISDN system through a local interface connected to a digital transmission medium called a digital pipe. There are several sizes of pipe available with varying capacities (i.e., bit rates), depending on customer need. For example, a residential cus- tomer may require only enough capacity to accommodate a telephone and a personal computer. However, an office complex may require a pipe with sufficient capacity to handle a large num- ber of digital telephones interconnected through an on-premise private branch exchange (PBX) or a large number of computers on a local area network (LAN). Figure 13 shows the ISDN user network, which illustrates the variety of network users and the need for more than one capacity pipe. A single residential telephone is at the low end of the ISDN demand curve, followed by a multiple-drop arrangement serving a telephone, a personal computer, and a home alarm system. Industrial complexes would be at the high end of the demand curve, as they require sufficient capacity to handle hundreds of telephones and several LANs.Although a pipe has a fixed capacity, the traffic on the pipe can be comprised of data from a dynamic variety of sources with varying signal types and bit rates that have been multiplexed into a single high-capacity pipe. Therefore, a cus- tomer can gain access to both circuit- and packet-switched services through the same Data-Link Protocols and Data Communications Networks 249
- 255. ISDN System Other applications Specialized- storage and information processing applications Single ISDN terminal Multiple ISDN terminals ISDN user-to-network interface PBX, LAN, private-line networks FIGURE 13 ISDN user network pipe. Because of the obvious complexity of ISDN, it requires a rather complex control sys- tem to facilitate multiplexing and demultiplexing data to provide the required services. 10-4 ISDN Objectives The key objectives of developing a worldwide ISDN system are the following: 1. System standardization. Ensure universal access to the network. 2. Achieving transparency. Allow customers to use a variety of protocols and appli- cations. 3. Separating functions. ISDN should not provide services that preclude competi- tiveness. 4. Variety of configurations. Provide private-line (leased) and switched services. 5. Addressing cost-related tariffs. ISDN service should be directly related to cost and independent of the nature of the data. 6. Migration. Provide a smooth transition while evolving. 7. Multiplexed support. Provide service to low-capacity personal subscribers as well as to large companies. 10-5 ISDN Architecture Figure 14 shows a block diagram of the architecture for ISDN functions. The ISDN net- work is designed to support an entirely new physical connector for the customer, a digital subscriber loop, and a variety of transmission services. A common physical is defined to provide a standard interface connection. A single in- terface will be used for telephones, computer terminals, and video equipment. Therefore, various protocols are provided that allow the exchange of control information between the customer’s device and the ISDN network. There are three basic types of ISDN channels: 1. B channel: 64 kbps 2. D channel: 16 kbps or 64 kbps 3. H channel: 384 kbps (H0), 1536 kbps (H11), or 1920 kbps (H12) ISDN standards specify that residential users of the network (i.e., the subscribers) be provided a basic access consisting of three full-duplex, time-division multiplexed digital channels, two operating at 64 kbps (designated the B channels, for bearer) and one at 16 kbps (designated the D channel, for data). The B and D bit rates were selected to be compatible with existing DS1–DS4 digital carrier systems. The D channel is used for carrying signaling Data-Link Protocols and Data Communications Networks 250
- 256. Frame-mode capabilities Packet switching Circuit switching ISDN Switch Network termination (NT) Customer premises Customer ISDN interface Digital subscriber loop to central office User-to-network signaling User-to-network signaling Terminal equipment (TE) ISDN Switch Customer site or service provider Nonswitched facilities Common channel signaling ISDN Network FIGURE 14 ISDN architecture information and for exchanging network control information. One B channel is used for dig- itally encoded voice and the other for applications such as data transmission, PCM-encoded digitized voice, and videotex. The 2B D service is sometimes called the basic rate inter- face (BRI). BRI systems require bandwidths that can accommodate two 64-kbps B channels and one 16-kbps D channel plus framing, synchronization, and other overhead bits for a to- tal bit rate of 192 kbps. The H channels are used to provide higher bit rates for special serv- ices such as fast facsimile, video, high-speed data, and high-quality audio. There is another service called the primary service, primary access, or primary rate interface (PRI) that will provide multiple 64-kbps channels intended to be used by the higher-volume subscribers to the network. In the United States, Canada, Japan, and Ko- rea, the primary rate interface consists of 23 64-kbps B channels and one 64-kbps D chan- nel (23B D) for a combined bit rate of 1.544 Mbps. In Europe, the primary rate inter- face uses 30 64-kbps B channels and one 64-kbps D channel for a combined bit rate of 2.048 Mbps. It is intended that ISDN provide a circuit-switched B channel with the existing tele- phone system; however, packet-switched B channels for data transmission at nonstandard rates would have to be created. The subscriber’s loop, as with the twisted-pair cable used with a common telephone, provides the physical signal path from the subscriber’s equipment to the ISDN central of- fice. The subscriber loop must be capable of supporting full-duplex digital transmission for both basic and primary data rates. Ideally, as the network grows, optical fiber cables will replace the metallic cables. Table 7 lists the services provided to ISDN subscribers. BC designates a circuit- switched B channel, BP designates a packet-switched B channel, and D designates a D channel. 10-6 ISDN System Connections and Interface Units ISDN subscriber units and interfaces are defined by their function and reference within the network. Figure 15 shows how users may be connected to an ISDN. As the figure shows, subscribers must access the network through one of two different types of entry devices: terminal equipment type 1 (TE1) or terminal equipment type 2 (TE2). TE1 equipment sup- ports standard ISDN interfaces and, therefore, requires no protocol translation. Data enter Data-Link Protocols and Data Communications Networks 251
- 257. ET TE1 ISDN terminal equipment S Customer NT12 Local loop Common carrier facilities Central office S S T U V TE2 Non-ISDN terminal equipment TA ISDN terminal adapter TE1 (Digital) ISDN telephone NT1 Subscriber line terminator LT NT2 Customer premises switching equipment R FIGURE 15 ISDN connections and reference points the network and are immediately configured into ISDN protocol format. TE2 equipment is classified as non-ISDN; thus, computer terminals are connected to the system through physical interfaces such as the RS-232 and host computers with X.25. Translation be- tween non-ISDN data protocol and ISDN protocol is performed in a device called a terminal adapter (TA), which converts the user’s data into the 64-kbps ISDN channel B or the 16-kbps channel D format and X.25 packets into ISDN packet formats. If any addi- tional signaling is required, it is added by the terminal adapter. The terminal adapters can also support traditional analog telephones and facsimile signals by using a 3.1-kHz audio service channel. The analog signals are digitized and put into ISDN format before enter- ing the network. User data at points designated as reference point S (system) are presently in ISDN for- mat and provide the 2B D data at 192 kbps. These reference points separate user terminal equipment from network-related system functions. Reference point T (terminal) locations correspond to a minimal ISDN network termination at the user’s location. These reference Data-Link Protocols and Data Communications Networks Table 7 ISDN Services Service Transmission Rate Channel Telephone 64 kbps BC System alarms 100 bps D Utility company metering 100 bps D Energy management 100 bps D Video 2.4–64 kbps BP Electronic mail 4.8–64 kbps BP Facsimile 4.8–64 kbps BC Slow-scan television 64 kbps BC 252
- 258. points separate the network provider’s equipment from the user’s equipment. Reference point R (rate) provides an interface between non-ISDN-compatible user equipment and the termi- nal adapters. Network termination 1 (NT1) provides the functions associated with the physical interface between the user and the common carrier and are designated by the letter T (these functions correspond to OSI layer 1). The NT1 is a boundary to the network and may be controlled by the ISDN provider. The NT1 performs line maintenance functions and supports multiple channels at the physical level (e.g., 2B D). Data from these chan- nels are time-division multiplexed together. Network terminal 2 devices are intelligent and can perform concentration and switching functions (functionally up through OSI level 3). NT2 terminations can also be used to terminate several S-point connections and provide local switching functions and two-wire-to-four-wire and four-wire-to-two-wire conversions. U-reference points refer to interfaces between the common carrier sub- scriber loop and the central office switch. A U loop is the media interface point between an NT1 and the central office. Network termination 1,2 (NT12) constitutes one piece of equipment that combines the functions of NT1 and NT2. U loops are terminated at the central office by a line termi- nation (LT) unit, which provides physical layer interface functions between the central of- fice and the loop lines. The LT unit is connected to an exchange termination (ET) at reference point V. An ET routes data to an outgoing channel or central office user. There are several types of transmission channels in addition to the B and D types de- scribed in the previous section. They include the following: HO channel. This interface supports multiple 384-kbps HO channels. These struc- tures are 3HO D and 4HO D for the 1.544-Mbps interface and 5HO D for the 2.048-Mbps interface. H11 channel. This interface consists of one 1.536-Mbps H11 channel (24 64-kbps channels). H12 channel. European version of H11 that uses 30 channels for a combined data rate of 1.92 Mbps. E channel. Packet switched using 64 kbps (similar to the standard D channel). 10-7 Broadband ISDN Broadband ISDN (BISDN) is defined by the ITU-T as a service that provides transmission channels capable of supporting transmission rates greater than the primary data rate. With BISDN, services requiring data rates of a magnitude beyond those provided by ISDN, such as video transmission, will become available. With the advent of BISDN, the original con- cept of ISDN is being referred to as narrowband ISDN. In 1988, the ITU-T first recommended as part of its I-series recommendations relating to BISDN: I.113, Vocabulary of terms for broadband aspects of ISDN, and I.121, Broadband aspects of ISDN. These two documents are a consensus concerning the aspects of the future of BISDN.They outline preliminary descriptions of future standards and development work. The new BISDN standards are based on the concept of an asynchronous transfer mode (ATM), which will incorporate optical fiber cable as the transmission medium for data transmission. The BISDN specifications set a maximum length of 1 km per cable length but are making provisions for repeated interface extensions. The expected data rates on the optical fiber cables will be either 11 Mbps, 155 Mbps, or 600 Mbps, depending on the specific application and the location of the fiber cable within the network. ITU-T classifies the services that could be provided by BISDN as interactive and dis- tribution services. Interactive services include those in which there is a two-way exchange Data-Link Protocols and Data Communications Networks 253
- 259. FIGURE 16 BISDN access of information (excluding control signaling) between two subscribers or between a sub- scriber and a service provider. Distribution services are those in which information trans- fer is primarily from service provider to subscriber. On the other hand, conversational ser- vices will provide a means for bidirectional end-to-end data transmission, in real time, between two subscribers or between a subscriber and a service provider. The authors of BISDN composed specifications that require the new services meet both existing ISDN interface specifications and the new BISDN needs. A standard ISDN terminal and a broadband terminal interface (BTI) will be serviced by the subscriber’s premise network (SPN), which will multiplex incoming data and transfer them to the broadband node. The broadband node is called a broadband network termination (BNT), which codes the data information into smaller packets used by the BISDN network. Data transmission within the BISDN network can be asymmetric (i.e., access on to and off of the network may be accomplished at different transmission rates, depending on system requirements). 10-7-1 BISDN configuration. Figure 16 shows how access to the BISDN net- work is accomplished. Each peripheral device is interfaced to the access node of a BISDN network through a broadband distant terminal (BDT). The BDT is responsible for the electrical-to-optical conversion, multiplexing of peripherals, and maintenance of the subscriber’s local system. Access nodes concentrate several BDTs into high-speed optical fiber lines directed through a feeder point into a service node. Most of the con- trol functions for system access are managed by the service node, such as call process- ing, administrative functions, and switching and maintenance functions. The functional modules are interconnected in a star configuration and include switching, administra- tive, gateway, and maintenance modules. The interconnection of the function modules is shown in Figure 17. The central control hub acts as the end user interface for con- trol signaling and data traffic maintenance. In essence, it oversees the operation of the modules. Subscriber terminals near the central office may bypass the access nodes entirely and be directly connected to the BISDN network through a service node. BISDN networks that use optical fiber cables can utilize much wider bandwidths and, consequently, have higher transmission rates and offer more channel-handling capacity than ISDN systems. Data-Link Protocols and Data Communications Networks 254
- 260. FIGURE 17 BISDN functional module interconnections 10-7-2 Broadband channel rates. The CCITT has published preliminary defini- tions of new broadband channel rates that will be added to the existing ISDN narrowband channel rates: 1. H21: 32.768 Mbps 2. H22: 43 Mbps to 45 Mbps 3. H4: 132 Mbps to 138.24 Mbps The H21 and H22 data rates are intended to be used for full-motion video transmis- sion for videoconferencing, video telephone, and video messaging. The H4 data rate is in- tended for bulk data transfer of text, facsimile, and enhanced video information. The H21 data rate is equivalent to 512 64-kbps channels. The H22 and H4 data rates must be multi- ples of the basic 64-kbps transmission rate. 11 ASYNCHRONOUS TRANSFER MODE Asynchronous transfer mode (ATM) is a relatively new data communications technology that uses a high-speed form of packet switching network for the transmission media. ATM was developed in 1988 by the ITU-T as part of the BISDN. ATM is one means by which data can enter and exit the BISDN network in an asynchronous (time-independent) fash- ion. ATM is intended to be a carrier service that provides an integrated, high-speed communications network for corporate private networks.ATM can handle all kinds of com- munications traffic, including voice, data, image, video, high-quality music, and multi- media. In addition, ATM can be used in both LAN and WAN network environments, pro- viding seamless internetworking between the two. Some experts claim that ATM may eventually replace both private leased T1 digital carrier systems and on-premise switching equipment. Conventional electronic switching (ESS) machines currently utilize a central proces- sor to establish switching paths and route traffic through a network. ATM switches, in con- trast, will include self-routing procedures where individual cells (short, fixed-length pack- ets of data) containing subscriber data will route their own way through the ATM switching Data-Link Protocols and Data Communications Networks 255
- 261. GFC 4 bits GFC—generic flow control field VPI —virtual path identifier VCI —virtual channel identifier PT —payload type identifier CLP —cell loss priority HEC—header error control VPI 4 bits 1 byte 1 byte 1 byte 1 byte 1 byte 6 bytes VPI 4 bits VCI 4 bits VCI 8 bits VCI 4 bits PT 3 bits CLP 1 bit HEC 8 bits Information field 48 bits FIGURE 19 ATM five-byte header field structure 5-byte header 48-byte information field 53-byte ATM cell FIGURE 18 ATM cell structure network in real time using their own address instead of relying on an external process to es- tablish the switching path. ATM uses virtual channels (VCs) and virtual paths (VPs) to route cells through a net- work. In essence, a virtual channel is merely a connection between a source and a destina- tion, which may entail establishing several ATM links between local switching centers. With ATM, all communications occur on the virtual channel, which preserves cell se- quence. On the other hand, a virtual path is a group of virtual channels connected between two points that could compromise several ATM links. ATM incorporates labeled channels that are transferable at fixed data rates anywhere from 16 kbps up to the maximum rate of the carrier system. Once data have entered the net- work, they are transferred into fixed time slots called cells. An ATM cell contains all the network information needed to relay individual cells from node to node over a preestab- lished ATM connection. Figure 18 shows the ATM cell structure, which is a fixed-length data packet only 53 bytes long, including a five-byte header and a 48-byte information field. Fixed-length cells provide the following advantages: 1. A uniform transmission time per cell ensures a more uniform transit-time charac- teristic for the network as a whole. 2. A short cell requires a shorter time to assemble and, thus, shorter delay character- istics for voice. 3. Short cells are easier to transfer over fixed-width processor buses, it is easier to buffer the data in link queues, and they require less processor logic. 11-1 ATM Header Field Figure 19 shows the five-byte ATM header field, which includes the following fields: generic flow control field, virtual path identifier, virtual channel identifier, payload type identifier, cell loss priority, and header error control. Generic flow control field (GFC). The GFC field uses the first four bits of the first byte of the header field. The GFC controls the flow of traffic across the user network inter- face (UNI) and into the network. Data-Link Protocols and Data Communications Networks 256
- 262. Virtual path identifier (VPI) and virtual channel identifier (VCI). The 24 bits immedi- ately following the GFC are used for the ATM address. Payload type identifier (PT). The first three bits of the second half of byte 4 specify the type of message (payload) in cell. With three bits, there are eight different types of pay- loads possible.At the present time, however, types 0 to 3 are used for identifying the type of user data, types 4 and 5 indicate management information, and types 6 and 7 are re- served for future use. Cell loss priority (CLP). The last bit of byte 4 is used to indicate whether a cell is eligi- ble to be discarded by the network during congested traffic periods. The CLP bit is set by the user or cleared by the user. If set, the network may discard the cell during times of heavy use. Header error control (HEC). The last byte of the header field is for error control and is used to detect and correct single-bit errors that occur in the header field only; the HEC does not serve as an entire cell check character. The value placed in the HEC is computed from the four previous bytes of the header field. The HEC provides some protection against the delivery of cells to the wrong destina- tion address. 11-1-1 ATM information field. The 48-byte information field is reserved for user data. Insertion of data into the information field of a cell is a function of the up- per half of layer 2 of the ISO-OSI seven-layer protocol hierarchy. This layer is specif- ically called the ATM Adaptation Layer (AAL). The AAL gives ATM the versatility necessary to facilitate, in a single format, a wide variety of different types of services ranging from continuous processes signals, such as voice transmission, to messages carrying highly fragmented bursts of data such as those produced from LANs. Because most user data occupy more than 48 bytes, the AAL divides information into 48-byte segments and places them into a series of segments. The five types of AALs are the following: 1. Constant bit rate (CBR). CBR information fields are designed to accommodate PCM-TDM traffic, which allows the ATM network to emulate voice or DSN services. 2. Variable bit rate (VBR) timing-sensitive services. This type of AAL is currently undefined; however, it is reserved for future data services requiring transfer of tim- ing information between terminal points as well as data (i.e., packet video). 3. Connection-oriented VBR data transfer. Type 3 information fields transfer VBR data such as impulsive data generated at irregular intervals between two sub- scribers over a preestablished data link. The data link is established by network signaling procedures that are very similar to those used by the public switched telephone network. This type of service is intended for large, long-duration data transfers, such as file transfers or file backups. 4. Connectionless VBR data transfer. This AAL type provides for transmission of VBR data that does not have a preestablished connection. Type 4 information fields are intended to be used for short, highly bursty types of transmissions, such as those generated from a LAN. 11-2 ATM Network Components Figure 20 shows an ATM network, which is comprised of three primary components: ATM endpoints, ATM switches, and transmission paths. Data-Link Protocols and Data Communications Networks 257
- 263. ATM Endpoint ATM Endpoint ATM Endpoint Transmission path ATM Endpoint ATM Endpoint Public ATM switch Public ATM switch Private ATM switch Private ATM switch Public ATM switch ATM Endpoint Transmission path Transmission paths Transmission paths Transmission paths Private ATM Network Private ATM Network Public ATM Network (common carrier) FIGURE 20 ATM network components 11-2-1 ATM endpoints. ATM endpoints are shown in Figure 21. As shown in the figure, endpoints are the source and destination of subscriber data and, therefore, they are sometimes called end systems. Endpoints can be connected directly to either a public or a private ATM switch. An ATM endpoint can be as simple as an ordinary personal computer equipped with an ATM network interface card. An ATM endpoint could also be a special- purpose network component that services several ordinary personal computers, such as an Ethernet LAN. 11-2-2 ATM switches. The primary function of an ATM switch is to route infor- mation from a source endpoint to a destination endpoint. ATM switches are sometimes called intermediate systems, as they are located between two endpoints. ATM switches fall into two general categories: public and private. Public ATM switches. A public ATM switch is simply a portion of a public service provider’s switching system where the service provider could be a local telephone company or a long-distance carrier, such as ATT. An ATM switch is sometimes called a network node. Private ATM switches. Private ATM switches are owned and maintained by a private company and sometimes called customer premise nodes. Private ATM switches are Data-Link Protocols and Data Communications Networks 258
- 264. ATM NIC Ethernet NIC ATM Switch ATM network ATM Switch Network software ATM NIC ATM Driver Network software ATM Endpoint ATM Endpoint ATM Endpoint Ethernet NIC Ethernet Driver ATM Switch Network software Ethernet NIC Ethernet Driver Ethernet LAN FIGURE 21 ATM endpoint implementations Data-Link Protocols and Data Communications Networks sold toATM customers by many of the same computer networking infrastructure ven- dors who provide ATM customers with network interface cards and connectivity de- vices, such as repeaters, hubs, bridges, switches, and routers. 11-2-3 Transmission paths. ATM switches and ATM endpoints are interconnected with physical communications paths called transmission paths. A transmission path can be any of the common transmission media, such as twisted-pair cable or optical fiber cable. 12 LOCAL AREA NETWORKS Studies have indicated that most (80%) of the communications among data terminals and other data equipment occurs within a relatively small local environment. A local area network (LAN) provides the most economical and effective means of handling lo- cal data communication needs. A LAN is typically a privately owned data communica- tions system in which the users share resources, including software. LANs provide two- way communications between a large variety of data communications terminals within a 259
- 265. Data-Link Protocols and Data Communications Networks PC Server Workstation Printer Scanner Workstation FIGURE 22 Typical local area network component configuration limited geographical area such as within the same room, building, or building complex. Most LANs link equipment that is within a few miles of each other. Figure 22 shows several personal computers (PCs) connected to a LAN to share com- mon resources such as a modem, printer, or server. The server may be a more powerful com- puter than the other PCs sharing the network, or it may simply have more disk storage space. The server “serves” information to the other PCs on the network in the form of soft- ware and data information files. A PC server is analogous to a mainframe computer except on a much smaller scale. LANs allow for a room full or more of computers to share common resources such as printers and modems. The average PC uses these devices only a small percentage of the time, so there is no need to dedicate individual printers and modems to each PC. To print a document or file, a PC simply sends the information over the network to the server. The server organizes and prioritizes the documents and then sends them, one document at a time, to the common usage printer. Meanwhile, the PCs are free to continue performing other useful tasks. When a PC needs a modem, the network establishes a virtual connec- tion between the modem and the PC. The network is transparent to the virtual connection, which allows the PC to communicate with the modem as if they were connected directly to each other. LANS allow people to send and receive messages and documents through the network much quicker than they could be sent through a paper mail system. Electronic mail (e-mail) is a communications system that allows users to send messages to each other through their computers. E-mail enables any PC on the network to send or receive information from any other PC on the network as long as the PCs and the server use the same or compatible soft- ware. E-mail can also be used to interconnect users on different networks in different cities, states, countries, or even continents. To send an e-mail message, a user at one PC sends its address and message along with the destination address to the server. The server effectively “relays” the message to the destination PC if they are subscribers to the same network. If the destination PC is busy or not available for whatever reason, the server stores the mes- sage and resends it later. The server is the only computer that has to keep track of the loca- tion and address of all the other PCs on the network. To send e-mail to subscribers of other networks, the server relays the message to the server on the destination user’s network, which in turn relays the mail to the destination PC. E-mail can be used to send text infor- mation (letters) as well as program files, graphics, audio, and even video. This is referred to as multimedia communications. 260
- 266. Data-Link Protocols and Data Communications Networks LANs are used extensively to interconnect a wide range of data services, including the following: Data terminals Data modems Laser printers Databases Graphic plotters Word processors Large-volume disk and tape storage devices Public switched telephone networks Facsimile machines Digital carrier systems (T carriers) Personal computers E-mail servers Mainframe computers 12-1 LAN System Considerations The capabilities of a LAN are established primarily by three factors: topology, transmission medium, and access control protocol. Together these three factors determine the type of data, rate of transmission, efficiency, and applications that a network can effectively support. 12-1-1 LAN topologies. The topology or physical architecture of a LAN identifies how the stations (terminals, printers, modems, and so on) are interconnected. The trans- mission media used with LANs include metallic twisted-wire pairs, coaxial cable, and optical fiber cables. Presently, most LANs use coaxial cable; however, optical fiber cable systems are being installed in many new networks. Fiber systems can operate at higher transmission bit rates and have a larger capacity to transfer information than coaxial cables. The most common LAN topologies are the star, bus, bus tree, and ring, which are il- lustrated in Figure 23. 12-1-2 Star topology. The preeminent feature of the star topology is that each sta- tion is radially linked to a central node through a direct point-to-point connection as shown in Figure 23a. With a star configuration, a transmission from one station enters the central node, where it is retransmitted on all the outgoing links. Therefore, although the circuit arrangement physically resembles a star, it is logically configured as a bus (i.e., transmis- sions from any station are received by all other stations). Central nodes offer a convenient location for system or station troubleshooting be- cause all traffic between outlying nodes must flow through the central node. The central node is sometimes referred to as central control, star coupler, or central switch and typi- cally is a computer. The star configuration is best adapted to applications where most of the communications occur between the central node and outlying nodes. The star arrangement is also well suited to systems where there is a large demand to communicate with only a few of the remote terminals. Time-sharing systems are generally configured with a star topol- ogy.A star configuration is also well suited for word processing and database management applications. Star couplers can be implemented either passively or actively. When passive couplers are used with a metallic transmission medium, transformers in the coupler provide an elec- tromagnetic linkage through the coupler, which passes incoming signals on to outgoing links. If optical fiber cables are used for the transmission media, coupling can be achieved by fusing fibers together. With active couplers, digital circuitry in the central node acts as a repeater. Incoming data are simply regenerated and repeated on to all outgoing lines. One disadvantage of a star topology is that the network is only as reliable as the cen- tral node. When the central node fails, the system fails. If one or more outlying nodes fail, however, the rest of the users can continue to use the remainder of the network. When fail- ure of any single entity within a network is critical to the point that it will disrupt service on the entire network, that entity is referred to as a critical resource. Thus, the central node in a star configuration is a critical resource. 261
- 267. Data-Link Protocols and Data Communications Networks 12-1-3 Bus topology. In essence, the bus topology is a multipoint or multidrop cir- cuit configuration where individual nodes are interconnected by a common, shared com- ing appropriate interfacing hardware, directly to a common linear transmission medium, generally referred to as a bus. In a bus configuration, network control is not centralized to a particular node. In fact, the most distinguishing feature of a bus LAN is that control is dis- tributed among all the nodes connected to the LAN. Data transmissions on a bus network are usually in the form of small packets containing user addresses and data. When one sta- tion desires to transmit data to another station, it monitors the bus first to determine if it is currently being used. If no other stations are communicating over the network (i.e., the net- work is clear), the monitoring station can commence to transmit its data. When one station begins transmitting, all other stations become receivers. Each receiver must monitor all transmission on the network and determine which are intended for them. When a station identifies its address on a received data message, it acts on it or it ignores that transmission. One advantage of a bus topology is that no special routing or circuit switching is required and, therefore, it is not necessary to store and retransmit messages intended for (a) (b) (c) (d) FIGURE 23 LAN Topologies: (a) star; (b) bus; (c) tree bus; (d) ring or loop munications channel as shown in Figure 23b.With the bus topology, all stations connect, us- 262
- 268. Data-Link Protocols and Data Communications Networks other nodes. This advantage eliminates a considerable amount of message identification overhead and processing time. However, with heavy-usage systems, there is a high likeli- hood that more than one station may desire to transmit at the same time. When transmissions from two or more stations occur simultaneously, a data collision occurs, disrupting data com- munications on the entire network. Obviously, a priority contention scheme is necessary to handle data collision. Such a priority scheme is called carrier sense, multiple access with collision detect (CSMA/CD), which is discussed in a later section of this chapter. Because network control is not centralized in a bus configuration, a node failure will not disrupt data flow on the entire LAN. The critical resource in this case is not a node but instead the bus itself. A failure anywhere along the bus opens the network and, depending on the ver- satility of the communications channel, may disrupt communication on the entire network. The addition of new nodes on a bus can sometimes be a problem because gaining ac- cess to the bus cable may be a cumbersome task, especially if it is enclosed within a wall, floor, or ceiling. One means of reducing installation problems is to add secondary buses to the primary communications channel. By branching off into other bases, a multiple bus structure called a tree bus is formed. Figure 23c shows a tree bus configuration. 12-1-4 Ring topology. With a ring topology, adjacent stations are interconnected by repeaters in a closed-loop configuration as shown in Figure 23d. Each node participates as a repeater between two adjacent links within the ring. The repeaters are relatively sim- ple devices capable of receiving data from one link and retransmitting them on a second link. Messages, usually in packet form, are propagated in the simplex mode (one-way only) from node to node around the ring until it has circled the entire loop and returned to the orig- inating node, where it is verified that the data in the returned message are identical to the data originally transmitted. Hence, the network configuration serves as an inherent error- detection mechanism. The destination station(s) can acknowledge reception of the data by setting or clearing appropriate bits within the control segment of the message packet. Pack- ets contain both source and destination address fields as well as additional network control information and user data. Each node examines incoming data packets, copying packets designated for them and acting as a repeater for all data packets by retransmitting them (bit by bit) to the next down-line repeater.A repeater should neither alter the content of received packets nor change the transmission rate. Virtuallyanyphysicaltransmissionmediumcanbeusedwiththeringtopology.Twisted- wirepairsofferlowcostbutseverelylimitedtransmissionrates.Coaxialcablesprovidegreater capacity than twisted-wire pairs at practically the same cost. The highest data rates, however, are achieved with optical fiber cables, except at a substantially higher installation cost. 12-2 LAN Transmission Formats Two transmission techniques or formats are used with LANs, baseband and broadband, to multiplex transmissions from a multitude of stations onto a single transmission medium. 12-2-1 Baseband transmission format. Baseband transmission formats are defined as transmission formats that use digital signaling. In addition, baseband formats use the trans- mission medium as a single-channel device. Only one station can transmit at a time, and all sta- tions must transmit and receive the same types of signals (encoding schemes, bit rates, and so on). Baseband transmission formats time-division multiplex signals onto the transmission medium. All stations can use the media but only one at a time. The entire frequency spectrum (bandwidth) is used by (or at least made available to) whichever station is presently transmit- ting. With a baseband format, transmissions are bidirectional.A signal inserted at any point on the transmission medium propagates in both directions to the ends, where it is absorbed. Dig- ital signaling requires a bus topology because digital signals cannot be easily propagated through the splitters and joiners necessary in a tree bus topology. Because of transmission line losses, baseband LANs are limited to a distance of no more than a couple miles. 263
- 269. Data-Link Protocols and Data Communications Networks Table 8 Baseband versus Broadband Transmission Formats Baseband Broadband Uses digital signaling Analog signaling requiring RF modems and amplifiers Entire bandwidth used by each transmission—no FDM possible, i.e., multiple data channels (video, FDM audio, data, etc.) Bidirectional Unidirectional Bus topology Bus or tree bus topology Maximum length approximately 1500 m Maximum length up to tens of kilometers Advantages Less expensive High capacity Simpler technology Multiple traffic types Easier and quicker to install More flexible circuit configurations, larger area covered Disadvantages Single channel RF modem and amplifiers required Limited capacity Complex installation and maintenance Grounding problems Double propagation delay Limited distance 12-2-2 Broadband transmission formats. Broadband transmission formats use the connecting media as a multichannel device. Each channel occupies a different fre- quency band within the total allocated bandwidth (i.e., frequency-division multiplexing). Consequently, each channel can contain different modulation and encoding schemes and operate at different transmission rates.A broadband network permits voice, digital data, and video to be transmitted simultaneously over the same transmission medium. However, broadband systems are unidirectional and require RF modems, amplifiers, and more com- plicated transceivers than baseband systems. For this reason, baseband systems are more prevalent. Circuit components used with broadband LANs easily facilitate splitting and joining operations; consequently, both bus and tree bus topologies are allowed. Broadband systems can span much greater distances than baseband systems. Distances of up to tens of miles are possible. The layout for a baseband system is much less complex than a broadband system and, therefore, easier and less expensive to implement. The primary disadvantages of baseband are its limited capacity and length. Broadband systems can carry a wide variety of different kinds of signals on a number of channels. By incorporating amplifiers, broadband can span much greater distances than baseband. Table 8 summarizes baseband and broadband transmission formats. 12-3 LAN Access Control Methodologies In a practical LAN, it is very likely that more than one user may wish to use the network media at any given time. For a medium to be shared by various users, a means of control- ling access is necessary. Media-sharing methods are known as access methodologies. Net- work access methodologies describe how users access the communications channel in a LAN. The first LANs were developed by computer manufacturers; they were expensive and worked only with certain types of computers with a limited number of software programs. LANs also required a high degree of technical knowledge and expertise to install and main- tain. In 1980, the IEEE, in an effort to resolve problems with LANs, formed the 802 Local Area Network Standards Committee. In 1983, the committee established several recom- mended standards for LANs. The two most prominent standards are IEEE Standard 802.3, which addresses an access method for bus topologies called carrier sense, multiple access with collision detection (CSMA/CD), and IEEE Standard 802.5, which describes an access method for ring topologies called token passing. 264
- 270. Data-Link Protocols and Data Communications Networks 12-3-1 Carrier sense, multiple access with collision detection. CSMA/CD is an ac- cess method used primarily with LANs configured in a bus topology. CSMA/CD uses the basic philosophy that, “If you have something to say, say it. If there’s a problem, we’ll work it out later.”With CSMA/CD, any station (node) can send a message to any other station (or stations) as long as the transmission medium is free of transmissions from other stations. Stations moni- tor (listen to) the line to determine if the line is busy. If a station has a message to transmit but the line is busy, it waits for an idle condition before transmitting its message. If two stations trans- mit at the same time, a collision occurs.When this happens, the station first sensing the collision sends a special jamming signal to all other stations on the network.All stations then cease trans- mitting (back off)and wait a random period of time before attempting a retransmission.The ran- dom delay time for each station is different and, therefore, allows for prioritizing the stations on the network. If successive collisions occur, the back-off period for each station is doubled. With CSMA/CD, stations must contend for the network. A station is not guaranteed access to the network. To detect the occurrence of a collision, a station must be capable of transmitting and receiving simultaneously. CSMA/CD is used by most LANs configured in a bus topology. Ethernet is an example of a LAN that uses CSMA/CD and is described later in this chapter. Another factor that could possibly cause collisions with CSMA/CD is propagation delay. Propagation delay is the time it takes a signal to travel from a source to a destination. Because of propagation delay, it is possible for the line to appear idle when, in fact, another station is transmitting a signal that has not yet reached the monitoring station. 12-3-2 Token passing. Token passing is a network access method used primarily with LANs configured in a ring topology using either baseband or broadband transmission formats. When using token passing access, nodes do not contend for the right to transmit data. With token passing, a specific packet of data, called a token, is circulated around the ring from station to station, always in the same direction. The token is generated by a des- ignated station known as the active monitor. Before a station is allowed to transmit, it must first possess the token. Each station, in turn, acquires the token and examines the data frame to determine if it is carrying a packet addressed to it. If the frame contains a packet with the receiving station’s address, it copies the packet into memory, appends any messages it has to send to the token, and then relinquishes the token by retransmitting all data packets and the token to the next node on the network. With token passing, each station has equal ac- cess to the transmission medium. As with CSMA/CD, each transmitted packet contains source and destination address fields. Successful delivery of a data frame is confirmed by the destination station by setting frame status flags, then forwarding the frame around the ring to the original transmitting station. The packet then is removed from the frame before transmitting the token. A token cannot be used twice, and there is a time limitation on how long a token can be held. This prevents one station from disrupting data transmissions on the network by holding the token until it has a packet to transmit. When a station does not possess the token, it can only receive and transfer other packets destined to other stations. Some 16-Mbps token ring networks use a modified form of token passing methodol- ogy where the token is relinquished as soon as a data frame has been transmitted instead of waiting until the transmitted data frame has been returned. This is known as an early token release mechanism. 13 ETHERNET Ethernet is a baseband transmission system designed in 1972 by Robert Metcalfe and David Boggs of the Xerox Palo Alto Research Center (PARC). Metcalfe, who later founded 3COM Corporation, and his colleagues at Xerox developed the first experimental Ethernet system to interconnect a Xerox Alto personal workstation to a graphical user interface. The 265
- 271. Data-Link Protocols and Data Communications Networks first experimental Ethernet system was later used to link Altos workstations to each other and to link the workstations to servers and laser printers. The signal clock for the experi- mental Ethernet interface was derived from the Alto’s system clock, which produced a data transmission rate of 2.94 Mbps. Metcalfe’s first Ethernet was called theAltoAloha Network; however, in 1973 Metcalfe changed the name to Ethernet to emphasize the point that the system could support any com- puter, not just Altos, and to stress the fact that the capabilities of his new network had evolved well beyond the original Aloha system. Metcalfe chose the name based on the word ether, meaning “air,” “atmosphere,” or “heavens,” as an indirect means of describing a vital feature of the system: the physical medium (i.e., a cable). The physical medium carries data bits to all stations in much the same way that luminiferous ether was once believed to transport electro- magnetic waves through space. In July 1976, Metcalfe and Boggs published a landmark paper titled “Ethernet: Dis- tributed Packet Switching for Local Computer.” On December 13, 1977, Xerox Corporation received patent number 4,063,220 titled “Multipoint Data Communications System with Collision Detection.” In 1979, Xerox joined forces with Intel and Digital Equipment Corpo- ration (DEC) in an attempt to make Ethernet an industry standard. In September 1980, the three companies jointly released the first version of the first Ethernet specification called the Ethernet Blue Book, DIX 1.0 (after the initials of the three companies), or Ethernet I. Ethernet I was replaced in November 1982 by the second version, called Ethernet II (DIX 2.0), which remains the current standard. In 1983, the 802 Working Group of the IEEE released their first standard for Ethernet technology. The formal title of the standard was IEEE 802.3 Carrier Sense Multiple Access with Collision Detection (CSMA/CD) Ac- cess Method and Physical Layer Specifications. The IEEE subsequently reworked several sections of the original standard, especially in the area of the frame format definition, and in 1985 they released the 802.3a standard, which was called thin Ethernet, cheapernet, or 10Base-2 Ethernet. In 1985, the IEEE also released the IEEE 802.3b 10Broad36 standard, which defined a transmission rate of 10 Mbps over a coaxial cable system. In 1987, two additional standards were released: IEEE 802.3d and IEEE 802.3e. The 802.3d standard defined the Fiber Optic Inter-Repeater Link (FOIRL) that used two fiber optic cables to extend the maximum distance between 10 Mbps repeaters to 1000 meters. The IEEE 802.3e standard defined a 1-Mbps standard based on twisted-pair cable, which was never widely accepted. In 1990, the IEEE introduced a major advance in Ethernet stan- dards: IEEE 802.3i. The 802.3i standard defined 10Base-T, which permitted a 10-Mbps transmission rate over simple category 3 unshielded twisted-pair (UTP) cable. The wide- spread use of UTP cabling in existing buildings created a high demand for 10Base-T tech- nology. 10Base-T also facilitated a star topology, which made it much easier to install, man- age, and troubleshoot. These advantages led to a vast expansion in the use of Ethernet. In 1993, the IEEE released the 802.3j standard for 10Base-F (FP, FB, and FL), which permitted attachment over longer distances (2000 meters) through two optical fiber cables. This standard updated and expanded the earlier FOIRL standard. In 1995, the IEEE improved the performance of Ethernet technology by a factor of 10 when it re- leased the 100-Mbps 802.3u 100Base-T standard. This version of Ethernet is commonly known as fast Ethernet. Fast Ethernet supported three media types: 100Base-TX, which operates over two pairs of category 5 twisted-pair cable; 100Base-T4, which operates over four pairs of category 3 twisted-pair cable; and 100Base-FX, which operates over two multimode fibers. In 1997, the IEEE released the 802.3x standard, which defined full-duplex Ethernet operation. Full-duplex Ethernet bypasses the normal CSMA/CD protocol and allows two stations to communicate over a point-to-point link, which effectively doubles the transfer rate by allowing each station to simultaneously transmit and receive separate data streams. In 1997, the IEEE also released the IEEE 802.3y 100Base-T2 standard for 100-Mbps op- eration over two pairs of category 3 balanced transmission line. 266
- 272. Data-Link Protocols and Data Communications Networks Table 9 Current IEEE Ethernet Standards Transmission Ethernet Transmission Maximum Rate System Medium Segment Length 10 Mbps 10Base-5 Coaxial cable (RG-8 or RG-11) 500 meters 10Base-2 Coaxial cable (RG-58) 185 meters 10Base-T UTP/STP category 3 or better 100 meters 10Broad-36 Coaxial cable (75 ohm) Varies 10Base-FL Optical fiber 2000 meters 10Base-FB Optical fiber 2000 meters 10Base-FP Optical fiber 2000 meters 100 Mbps 100Base-T UTP/STP category 5 or better 100 meters 100Base-TX UTP/STP category 5 or better 100 meters 100Base-FX Optical fiber 400–2000 meters 100Base-T4 UTP/STP category 5 or better 100 meters 1000 Mbps 1000Base-LX Long-wave optical fiber Varies 1000Base-SX Short-wave optical fiber Varies 1000Base-CX Short copper jumper Varies 1000Base-T UTP/STP category 5 or better Varies In 1998, IEEE once again improved the performance of Ethernet technology by a fac- tor of 10 when it released the 1-Gbps 802.3z 1000Base-X standard, which is commonly called gigabit Ethernet. Gigabit Ethernet supports three media types: 1000Base-SX, which operates with an 850-nm laser over multimode fiber; 1000Base-LX, which operates with a 1300-nm laser over single and multimode fiber; and 1000Base-CX, which operates over short-haul copper-shielded twisted-pair (STP) cable. In 1998, the IEEE also released the 802.3ac standard, which defines extensions to support virtual LAN (VLAN) tagging on Ethernet networks. In 1999, the release of the 802.3ab 1000Base-T standard defined 1-Gbps operation over four pairs of category 5 UTP cabling. The topology of choice for Ethernet LANs is either a linear bus or a star, and all Eth- ernet systems employ carrier sense, multiple access with collision detect (CSMA/CD) for the accessing method. 13-1 IEEE Ethernet Standard Notation To distinguish the various implementations of Ethernet available, the IEEE 802.3 commit- tee has developed a concise notation format that contains information about the Ethernet system, including such items as bit rate, transmission mode, transmission medium, and seg- ment length. The IEEE 802.3 format is data rate in Mbpstransmission modemaximum segment length in hundreds of meters or data rate in Mbpstransmission modetransmission media The transmission rates specified for Ethernet are 10 Mbps, 100 Mbps, and 1 Gbps. There are only two transmission modes: baseband (base) or broadband (broad). The segment length can vary, depending on the type of transmission medium, which could be coaxial ca- ble (no designation), twisted-pair cable (T), or optical fiber (F). For example, the notation 10Base-5 means 10-Mbps transmission rate, baseband mode of transmission, with a maxi- mum segment length of 500 meters. The notation 100Base-T specifies 100-Mbps trans- mission rate, baseband mode of transmission, with a twisted-pair transmission medium. The notation 100Base-F means 100-Mbps transmission rate, baseband transmission mode, with an optical fiber transmission medium. The IEEE currently supports nine 10-Mbps standards, six 100-Mbps standards, and five 1-Gbps standards. Table 9 lists several of the more common types of Ethernet, their ca- bling options, distances supported, and topology. 267
- 273. Data-Link Protocols and Data Communications Networks Station Repeater 1 Repeater 4 Repeater 2 Repeater 3 Segment 1 500 meters Segment 5 500 meters Segment 3 (no stations) 500 meters 500 meters Segment 2 (no stations) Station Station Segment 4 500 meters Station Station Station Station FIGURE 24 10 Mbps 4-3 Ethernet configuration 13-2 10-Mbps Ethernet Figure 24 shows the physical layout for a 10Base-5 Ethernet system. The maximum number of cable segments supported with 10Base-5 Ethernet is five, interconnected with four repeaters or hubs. However, only three of the segments can be populated with nodes (computers). This is called the 5-4-3 rule: five segments joined by four repeaters, but only three segments can be populated. The maximum segment length for 10Base-5 is 500 meters. Imposing maximum segment lengths are required for the CSMA/CD to operate properly. The limitations take into account Ethernet frame size, velocity of propagation on a given transmission medium, and re- peater delay time to ensure collisions that occur on the network are detected. On 10Base-5 Ethernet, the maximum segment length is 500 meters with a maximum of five segments. Therefore, the maximum distance between any two nodes (computers) is 5 500 2500 meters. The worst-case scenario for collision detection is when the station at one end of the network completes a transmission at the same instant the station at the far end of the network begins a transmission. In this case, the station that transmitted first would not know that a collision had occurred. To prevent this from happening, minimum frame lengths are imposed on Ethernet. The minimum frame length for 10Base-5 is computed as follows. The velocity of prop- agation along the cable is assumed to be approximately two-thirds the speed of light, or vp 2 108 ms vp ¢ 2 3 ≤13 108 ms2 vp 2 3 vc 268
- 274. Data-Link Protocols and Data Communications Networks Thus, the length of a bit along a cable for a bit rate of 10 Mbps is and the maximum number of bits on a cable with a maximum length of 2500 meters is Therefore, the maximum time for a bit to propagate end to end is and the round-trip delay equals 2 12.5 μs 25 μs Therefore, the minimum length of an Ethernet message for a 10-Mbps transmission rate is where the time of a bit (tb 1/bit rate or 1/10 Mbps 0.1 μs). However, the minimum num- ber of bits is doubled and rounded up to 512 bits (64 eight-bit bytes). 10Base-5 is the original Ethernet that specifies a thick 50-Ω double-shielded RG-11 coaxial cable for the transmission medium. Hence, this version is sometimes called thicknet or thick Ethernet. Because of its inflexible nature, 10Base-5 is sometimes called frozen yellow garden hose. 10Base-5 Ethernet uses a bus topology with an external device called a media access unit (MAU) to connect terminals to the cable.The MAU is sometimes called a vampire tap because it connects to the cable by simply puncturing the cable with a sharp prong that ex- tends into the cable until it makes contact with the center conductor. Each connection is called a tap, and the cable that connects the MAU to its terminal is called an attachment unit inter- face (AUI) or sometimes simply a drop.Within each MAU, a digital transceiver transfers elec- trical signals between the drop and the coaxial transmission medium. 10Base-5 supports a maximumof100nodespersegment.Repeatersarecountedasnodes;therefore,themaximum capacity of a 10Base-5 Ethernet is 297 nodes.With 10Base-5, unused taps must be terminated in a 50-Ω resistive load. A drop left unterminated or any break in the cable will cause total LAN failure. 13-2-1 10Base-2 Ethernet. 10Base-5 Ethernet uses a 50-ΩRG-11 coaxial cable, which is thick enough to give it high noise immunity, thus making it well suited to labora- tory and industrial applications. The RG-11 cable, however, is expensive to install. Conse- quently, the initial costs of implementing a 10Base-5 Ethernet system are too high for many small businesses. In an effort to reduce the cost, International Computer Ltd, Hewlett- Packard, and 3COM Corporation developed an Ethernet variation that uses thinner, less ex- pensive 50-ΩRG-58 coaxial cable. RG-58 is less expensive to purchase and install than RG- 11. In 1985, the IEEE 802.3 Standards Committee adopted a new version of Ethernet and gave it the name 10Base-2, which is sometimes called cheapernet or thinwire Ethernet. 10Base-2 Ethernet uses a bus topology and allows a maximum of five segments; how- ever, only three can be populated. Each segment has a maximum length of 185 meters with no more than 30 nodes per segment. This limits the capacity of a 10Base-2 network to 96 nodes. 10Base-2 eliminates the MAU, as the digital transceiver is located inside the termi- nal and a simple BNC-T connector connects the network interface card (NIC) directly to the coaxial cable. This eliminates the expensive cable and the need to tap or drill into it. round-trip delay bit time 25 μs 0.1 μs 250 bits 2500 m 2 108 ms 12.5 μs 2500 m 20 mb 125 bits bit length 2 108 ms 10 mbps 20 metersbit 269
- 275. Data-Link Protocols and Data Communications Networks With 10Base-2 Ethernet, unused taps must be terminated in a 50-Ω resistive load and a drop left unterminated, or any break in the cable will cause total LAN failure. 13-2-2 10Base-T Ethernet. 10Base-T Ethernet is the most popular 10-Mbps Eth- ernet commonly used with PC-based LAN environments utilizing a star topology. Because stations can be connected to a network hub through an internal transceiver, there is no need for anAUI. The “T” indicates unshielded twisted-pair cable. 10Base-T was developed to al- low Ethernet to utilize existing voice-grade telephone wiring to carry Ethernet signals. Standard modular RJ-45 telephone jacks and four-pair UTP telephone wire are specified in the standard for interconnecting nodes directly to the LAN without an AUI. The RJ-45 con- nector plugs direction into the network interface card in the PC. 10Base-T operates at a transmission rate of 10 Mbps and uses CSMA/CD; however, it uses a multiport hub at the center of network to interconnect devices. This essentially converts each segment to a point- to-point connection into the LAN. The maximum segment length is 100 meters with no more than two nodes on each segment. Nodes are added to the network through a port on the hub. When a node is turned on, its transceiver sends a DC current over the twisted-pair cable to the hub. The hub senses the current and enables the port, thus connecting the node to the network. The port remains con- nected as long as the node continues to supply DC current to the hub. If the node is turned off or if an open- or short-circuit condition occurs in the cable between the node and the hub, DC current stops flowing, and the hub disconnects the port from the network, allow- ing the remainder of the LAN to continue operating status quo. With 10Base-T Ethernet, a cable break affects only the nodes on that segment. 13-2-3 10Base-FL Ethernet. 10Base-FL (fiber link) is the most common 10-Mbps Ethernet that uses optical fiber for the transmission medium. 10Base-FL is arranged in a star topology where stations are connected to the network through an external AUI cable and an external transceiver called a fiber-optic MAU. The transceiver is connected to the hub with two pairs of optical fiber cable. The cable specified is graded-index multimode ca- ble with a 62.5-μm-diameter core. 13-3 100-Mbps Ethernet Over the past few years, it has become quite common for bandwidth-starved LANs to up- grade 10Base-T Ethernet LANs to 100Base-T (sometimes called fast Ethernet). The 100Base-T Ethernet includes a family of fast Ethernet standards offering 100-Mbps data transmission rates using CSMA/CD access methodology. 100-Mbps Ethernet installations do not have the same design rules as 10-Mbps Ethernet. 10-Mbps Ethernet allows several connections between hubs within the same segment (collision domain). 100-Mbps Ether- net does not allow this flexibility. Essentially, the hub must be connected to an internet- working device, such as a switch or a router. This is called the 2-1 rule—two hubs mini- mum for each switch. The reason for this requirement is for collision detection within a domain. The transmission rate increased by a factor of 10; therefore, frame size, cable prop- agation, and hub delay are more critical. IEEE standard 802.3u details operation of the 100Base-T network. There are three media-specific physical layer standards for 100Base-TX: 100Base-T, 100Base-T4, and 100Base-FX. 13-3-1 100Base-TX Ethernet. 100Base-TX Ethernet is the most common of the 100- Mbps Ethernet standards and the system with the most technology available. 100Base-TX specifies a 100-Mbps data transmission rate over two pairs of category 5 UTP or STP cables with a maximum segment length of 100 meters. 100Base-TX uses a physical star topology (half duplex) or bus (full duplex) with the same media access method (CSMA/CD) and frame structures as 10Base-T; however, 100Base-TX requires a hub port and NIC, both of which 270
- 276. Data-Link Protocols and Data Communications Networks must be 100Base-TX compliant. 100Base-TX can operate full duplex in certain situations, such as from a switch to a server. 13-3-2 100Base-T4 Ethernet. 100Base-T4 is a physical layer standard specifying 100-Mbps data rates using two pairs of category 3, 4, or 5 UTP or STP cable. 100Base-T4 was devised to allow installations that do not comply with category 5 UTP cabling specifi- cations. 100Base-T4 will operate using category 3 UTP installation or better; however, there are some significant differences in the signaling. 13-3-3 100Base-FX Ethernet. 100Base-FX is a physical layer standard specifying 100-Mbps data rates over two optical fiber cables using a physical star topology. The logi- cal topology for 100Base-FX can be either a star or a bus. 100Base-FX is often used to in- terconnect 100Base-TX LANs to a switch or router. 100Base-FX uses a duplex optical fiber connection with multimode cable that supports a variety of distances, depending on cir- cumstances. 13-4 1000-Mbps Ethernet One-gigabit Ethernet (1 GbE) is the latest implementation of Ethernet that operates at a trans- mission rate of one billion bits per second and higher. The IEEE 802.3z Working Group is currently preparing standards for implementing gigabit Ethernet. Early deployments of giga- bit Ethernet were used to interconnect 100-Mbps and gigabit Ethernet switches, and gigabit Ethernet is used to provide a fat pipe for high-density backbone connectivity. Gigabit Ether- net can use one of two approaches to medium access: half-duplex mode using CSMA/CD or full-duplex mode, where there is no need for multiple accessing. Gigabit Ethernet can be generally categorized as either two-wire 1000Base-X or four-wire 1000Base-T. Two-wire gigabit Ethernet can be either 1000Base-SX for short- wave optical fiber, 1000Base-LX for long-wave optical fiber, or 1000Base-CX for short copper jumpers. The four-wire version of gigabit Ethernet is 1000Base-T. 1000Base- SX and 1000Base-LX use two optical fiber cables where the only difference between them is the wavelength (color) of the light waves propagated through the cable. 1000Base-T Ethernet was designed to be used with four twisted pairs of Category 5 UTP cables. 13-5 Ethernet Frame Formats Over the years, four different Ethernet frame formats have emerged where network envi- ronment dictates which format is implemented for a particular configuration. The four for- mats are the following: Ethernet II. The original format used with DIX. IEEE 802.3. The first generation of the IEEE Standards Committee, often referred to as a raw IEEE 802.3 frame. Novell was the only software vendor to use this format. IEEE 802.3 with 802.2 LLC. Provides support for IEEE 802.2 LLC. IEEE 802.3 with SNAP similar to IEEE 802.3 but provides backward compatibility for 802.2 to Ethernet II formats and protocols. Ethernet II and IEEE 802.3 are the two most popular frame formats used with Ether- net. Although they are sometimes thought of as the same thing, in actuality Ethernet II and IEEE 802.3 are not identical, although the term Ethernet is generally used to refer to any IEEE 802.3-compliant network. Ethernet II and IEEE 802.3 both specify that data be trans- mitted from one station to another in groups of data called frames. 13-5-1 Ethernet II frame format. The frame format for Ethernet II is shown in Figure 25 and is comprised of a preamble, start frame delimiter, destination address, source address, type field, data field, and frame check sequence field. 271
- 277. Data-Link Protocols and Data Communications Networks Preamble 7 bytes Frame check field 4 bytes Destination address 6 bytes Start frame delimiter (SFD) 1 byte Source address 6 bytes Type field 2 bytes Data field 46 to 1500 bytes 00 00 00 4A 72 4C 1 0 1 0 1 0 1 1 MAC address of source 02 60 8C 4A 72 4C 32 pairs of alternating 1/0 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 MAC address of destination 32-Bit CRC character Network operating system, data, and IP PDU (0600 hex or higher) Indicates type of higher layer protocol contained in data field FIGURE 25 Ethernet II frame format Preamble. The preamble consists of eight bytes (64 bits) of alternating 1s and 0s. The purpose of the preamble is to establish clock synchronization. The last two bits of the pre- amble are reserved for the start frame delimiter. Start frame delimiter. The start frame delimiter is simply a series of two logic 1s ap- pended to the end of the preamble, whose purpose is to mark the end of the preamble and the beginning of the data frame. Destination address. The source and destination addresses and the field type make up the frame header. The destination address consists of six bytes (48 bits) and is the address of the node or nodes that have been designated to receive the frame. The address can be a unique, group, or broadcast address and is determined by the following bit combinations: bit 0 0 If bit 0 is a 0, the address is interpreted as a unique address in- tended for only one station. bit 0 1 If bit 0 is a 1, the address is interpreted as a multicast (group) address. All stations that have been preassigned with this group address will accept the frame. bit 0–47 If all bits in the destination field are 1s, this identifies a broad- cast address, and all nodes have been identified as receivers of this frame. Source address. The source address consists of six bytes (48 bits) that correspond to the address of the station sending the frame. Type field. Ethernet does not use the 16-bit type field. It is placed in the frame so it can be used for higher layers of the OSI protocol hierarchy. Data field. The data field contains the information and can be between 46 bytes and 1500 bytes long. The data field is transparent. Data-link control characters and zero-bit stuffing are not used. Transparency is achieved by counting back from the FCS character. Frame check sequence field. The CRC field contains 32 bits for error detection and is computed from the header and data fields. 13-5-2 IEEE 802.3 frame format. The frame format for IEEE 802.3 is shown in Figure 26 and consists of the following: Preamble. The preamble consists of seven bytes to establish clock synchronization. The last byte of the preamble is used for the start frame delimiter. Start frame delimiter. The start frame delimiter is simply a series of two logic 1s ap- pended to the end of the preamble, whose purpose is to mark the end of the preamble and the beginning of the data frame. 272
- 278. Data-Link Protocols and Data Communications Networks Preamble 7 bytes Frame check field 4 bytes Destination address 6 bytes Start frame delimiter (SFD) 1 byte Source address 6 bytes Length field 2 bytes Data field 46 to 1500 bytes 00 00 00 4A 72 4C 1 0 1 0 1 0 1 1 MAC address of source 02 60 8C 4A 72 4C 32 pairs of alternating 1/0 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 MAC address of destination 32-Bit CRC character IEEE 802.2 LLC, data, and IP PDU (002E to 05DC) Specifies the length of the data field FIGURE 26 IEEE 802.3 frame format Destination and source addresses. The destination and source addresses are defined the same as with Ethernet II. Length field. The two-byte length field in the IEEE 802.3 frame replaces the type field in the Ethernet frame. The length field indicates the length of the variable-length logical link control (LLC) data field, which contains all upper-layered embedded protocols. Logical link control (LLC). The LLC field contains the information and can be be- tween 46 bytes and 1500 bytes long. The LLC field defined in IEEE 802.3 is identical to the LLC field defined for token ring networks. Frame check sequence field. The CRC field is defined the same as with Ethernet II. End-of-frame delimiter. The end-of-frame delimiter is a period of time (9.6 μs) in which no bits are transmitted. With Manchester encoding, a void in transitions longer than one-bit time indicates the end of the frame. QUESTIONS 1. Define data-link protocol. 2. What is meant by a primary station? Secondary station? 3. What is a master station? Slave station? 4. List and describe the three data-link protocol functions. 5. Briefly describe the ENQ/ACK line discipline. 6. Briefly describe the poll/select line discipline. 7. Briefly describe the stop-and-wait method of flow control. 8. Briefly describe the sliding window method of flow control. 9. What is the difference between character- and bit-oriented protocols? 10. Describe the difference between asynchronous and synchronous protocols. 11. Briefly describe how the XMODEM protocol works. 12. Why is IBM’s 3270 protocol called “bisync”? 13. Briefly describe the polling sequence for BSC, including the difference between a general and specific poll. 14. Briefly describe the selection sequence for BSC. 15. How does BSC achieve transparency? 16. What is the difference between a command and a response with SDLC? 273
- 279. Data-Link Protocols and Data Communications Networks 17. What are the three transmission states used with SDLC? 18. What are the five fields used with SDLC? 19. What is the delimiting sequence used with SDLC? 20. What are the three frame formats used with SDLC? 21. What are the purposes of the ns and nr bit sequences? 22. What is the difference between P and F bits? 23. With SDLC, which frame types can contain an information field? 24. With SDLC, which frame types can be used for error correction? 25. What SDLC command/response is used for reporting procedural errors? 26. When is the configure command/response used with SDLC? 27. What is the go-ahead sequence? The turnaround sequence? 28. What is the transparency mechanism used with SDLC? 29. What supervisory condition exists with HDLC that is not included in SDLC? 30. What are the transparency mechanism and delimiting sequence for HDLC? 31. Briefly describe invert-on-zero encoding. 32. List and describe the HDLC operational modes. 33. Briefly describe the layout for a public switched data network. 34. What is a value-added network? 35. Briefly describe circuit, message, and packet switching. 36. What is a transactional switch? A transparent switch? 37. Explain the following terms: permanent virtual circuit, virtual call, and datagram. 38. Briefly describe an X.25 call request packet. 39. Briefly describe an X.25 data transfer packet. 40. Define ISDN. 41. List and describe the principles of ISDN. 42. List and describe the evolution of ISDN. 43. Describe the conceptual view of ISDN and what is meant by the term digital pipe. 44. List the objectives of ISDN. 45. Briefly describe the architecture of ISDN. 46. List and describe the ISDN system connections and interface units. 47. Briefly describe BISDN. 48. Briefly describe asynchronous transfer mode. 49. Describe the differences between virtual channels and virtual paths. 50. Briefly describe the ATM header field; ATM information field. 51. Describe the following ATM network components: ATM endpoints, ATM switches, ATM trans- mission paths. 52. Briefly describe a local area network. 53. List and describe the most common LAN topologies. 54. Describe the following LAN transmission formats: baseband and broadband. 55. Describe the two most common LAN access methodologies. 56. Briefly describe the history of Ethernet. 57. Describe the Ethernet standard notation. 58. List and briefly describe the 10-Mbps Ethernet systems. 59. List and briefly describe the 100-Mbps Ethernet systems. 60. List and briefly describe the 1000-Mbps Ethernet systems. 61. Describe the two most common Ethernet frame formats. 274
- 280. Data-Link Protocols and Data Communications Networks PROBLEMS 1. Determine the hex code for the control field in an SDLC frame for the following conditions: in- formation frame, poll, transmitting frame 4, and confirming reception of frames 2, 3, and 4. 2. Determine the hex code for the control field in an SDLC frame for the following conditions: su- pervisory frame, ready to receive, final, and confirming reception of frames 6, 7, and 0. 3. Insert 0s into the following SDLC data stream: 111 001 000 011 111 111 100 111 110 100 111 101 011 111 111 111 001 011 4. Delete 0s from the following SDLC data stream: 010 111 110 100 011 011 111 011 101 110 101 111 101 011 100 011 111 00 5. Sketch the NRZI waveform for the following data stream (start with a high condition): 1 0 0 1 1 1 0 0 1 0 1 0 6. Determine the hex code for the control field in an SDLC frame for the following conditions: in- formation frame, not a poll, transmitting frame number 5, and confirming the reception of frames 0, 1, 2, and 3. 7. Determine the hex code for the control field in an SDLC frame for the following conditions: su- pervisory frame, not ready to receive, not a final, and confirming reception of frames 7, 0, 1, and 2. 8. Insert 0s into the following SDLC data stream: 0110111111101100001111100101110001011111111011111001 9. Delete 0s from the following SDLC data stream: 0010111110011111011111011000100011111011101011000101 10. Sketch the NRZI levels for the following data stream (start with a high condition): 1 1 0 1 0 0 0 1 1 0 1 ANSWERS TO SELECTED PROBLEMS 1. B8 hex 3. 4 inserted zeros 5. 8A hex 7. 65 hex 9. 4 deleted zeros 275
- 281. 276
- 282. Digital Transmission CHAPTER OUTLINE 1 Introduction 9 Companding 2 Pulse Modulation 10 Vocoders 3 PCM 11 PCM Line Speed 4 PCM Sampling 12 Delta Modulation PCM 5 Signal-to-Quantization Noise Ratio 13 Adaptive Delta Modulation PCM 6 Linear versus Nonlinear PCM Codes 14 Differential PCM 7 Idle Channel Noise 15 Pulse Transmission 8 Coding Methods 16 Signal Power in Binary Digital Signals OBJECTIVES ■ Define digital transmission ■ List and describe the advantages and disadvantages of digital transmission ■ Briefly describe pulse width modulation, pulse position modulation, and pulse amplitude modulation ■ Define and describe pulse code modulation ■ Explain flat-top and natural sampling ■ Describe the Nyquist sampling theorem ■ Describe folded binary codes ■ Define and explain dynamic range ■ Explain PCM coding efficiency ■ Describe signal-to-quantization noise ratio ■ Explain the difference between linear and nonlinear PCM codes ■ Describe idle channel noise ■ Explain several common coding methods ■ Define companding and explain analog and digital companding ■ Define digital compression From Chapter 6 of Advanced Electronic Communications Systems, Sixth Edition. Wayne Tomasi. Copyright © 2004 by Pearson Education, Inc. Published by Pearson Prentice Hall. All rights reserved. 277
- 283. ■ Describe vocoders ■ Explain how to determine PCM line speed ■ Describe delta modulation PCM ■ Describe adaptive delta modulation ■ Define and describe differential pulse code modulation ■ Describe the composition of digital pulses ■ Explain intersymbol interference ■ Explain eye patterns ■ Explain the signal power distribution in binary digital signals 1 INTRODUCTION As stated previously, digital transmission is the transmittal of digital signals between two or more points in a communications system. The signals can be binary or any other form of discrete-level digital pulses. The original source information may be in digital form, or it could be analog signals that have been converted to digital pulses prior to transmission and converted back to analog signals in the receiver. With digital transmission systems, a phys- ical facility, such as a pair of wires, coaxial cable, or an optical fiber cable, is required to interconnect the various points within the system. The pulses are contained in and propa- gate down the cable. Digital pulses cannot be propagated through a wireless transmission system, such as Earth’s atmosphere or free space (vacuum). ATT developed the first digital transmission system for the purpose of carrying dig- itally encoded analog signals, such as the human voice, over metallic wire cables between telephone offices. Today, digital transmission systems are used to carry not only digitally encoded voice and video signals but also digital source information directly between com- puters and computer networks. Digital transmission systems use both metallic and optical fiber cables for their transmission medium. 1-1 Advantages of Digital Transmission The primary advantage of digital transmission over analog transmission is noise immunity. Digital signals are inherently less susceptible than analog signals to interference caused by noise because with digital signals it is not necessary to evaluate the precise amplitude, fre- quency, or phase to ascertain its logic condition. Instead, pulses are evaluated during a pre- cise time interval, and a simple determination is made whether the pulse is above or below a prescribed reference level. Digital signals are also better suited than analog signals for processing and combin- ing using a technique called multiplexing. Digital signal processing (DSP) is the process- ing of analog signals using digital methods and includes bandlimiting the signal with fil- ters, amplitude equalization, and phase shifting. It is much simpler to store digital signals than analog signals, and the transmission rate of digital signals can be easily changed to adapt to different environments and to interface with different types of equipment. In addition, digital transmission systems are more resistant to analog systems to ad- ditive noise because they use signal regeneration rather than signal amplification. Noise produced in electronic circuits is additive (i.e., it accumulates); therefore, the signal-to- noise ratio deteriorates each time an analog signal is amplified. Consequently, the number of circuits the signal must pass through limits the total distance analog signals can be trans- ported. However, digital regenerators sample noisy signals and then reproduce an entirely new digital signal with the same signal-to-noise ratio as the original transmitted signal. Therefore, digital signals can be transported longer distances than analog signals. Finally, digital signals are simpler to measure and evaluate than analog signals. Therefore, it is easier to compare the error performance of one digital system to another dig- ital system. Also, with digital signals, transmission errors can be detected and corrected more easily and more accurately than is possible with analog signals. Digital Transmission 278
- 284. 1-2 Disadvantages of Digital Transmission The transmission of digitally encoded analog signals requires significantly more bandwidth than simply transmitting the original analog signal. Bandwidth is one of the most important aspects of any communications system because it is costly and limited. Also, analog signals must be converted to digital pulses prior to transmission and con- verted back to their original analog form at the receiver, thus necessitating additional en- coding and decoding circuitry. In addition, digital transmission requires precise time syn- chronization between the clocks in the transmitters and receivers. Finally, digital transmission systems are incompatible with older analog transmission systems. 2 PULSE MODULATION Pulse modulation consists essentially of sampling analog information signals and then con- verting those samples into discrete pulses and transporting the pulses from a source to a des- tination over a physical transmission medium. The four predominant methods of pulse modulation include pulse width modulation (PWM), pulse position modulation (PPM), pulse amplitude modulation (PAM), and pulse code modulation (PCM). PWM is sometimes called pulse duration modulation (PDM) or pulse length modu- lation (PLM), as the width (active portion of the duty cycle) of a constant amplitude pulse is varied proportional to the amplitude of the analog signal at the time the signal is sampled. PWM is shown in Figure 1c. As the figure shows, the amplitude of sample 1 is lower than the amplitude of sample 2. Thus, pulse 1 is narrower than pulse 2. The maximum analog signal amplitude produces the widest pulse, and the minimum analog signal amplitude pro- duces the narrowest pulse. Note, however, that all pulses have the same amplitude. With PPM, the position of a constant-width pulse within a prescribed time slot is var- ied according to the amplitude of the sample of the analog signal. PPM is shown in Figure 1d. As the figure shows, the higher the amplitude of the sample, the farther to the right the pulse is positioned within the prescribed time slot. The highest amplitude sample produces a pulse to the far right, and the lowest amplitude sample produces a pulse to the far left. With PAM, the amplitude of a constant width, constant-position pulse is varied ac- cording to the amplitude of the sample of the analog signal. PAM is shown in Figure 1e, where it can be seen that the amplitude of a pulse coincides with the amplitude of the ana- log signal. PAM waveforms resemble the original analog signal more than the waveforms for PWM or PPM. With PCM, the analog signal is sampled and then converted to a serial n-bit binary code for transmission. Each code has the same number of bits and requires the same length of time for transmission. PCM is shown in Figure 1f. PAM is used as an intermediate form of modulation with PSK, QAM, and PCM, al- though it is seldom used by itself. PWM and PPM are used in special-purpose communi- cations systems mainly for the military but are seldom used for commercial digital trans- mission systems. PCM is by far the most prevalent form of pulse modulation and, consequently, will be discussed in more detail in subsequent sections of this chapter. 3 PCM Alex H. Reeves is credited with inventing PCM in 1937 while working forATT at its Paris laboratories. Although the merits of PCM were recognized early in its development, it was not until the mid-1960s, with the advent of solid-state electronics, that PCM became preva- lent. In the United States today, PCM is the preferred method of communications within the public switched telephone network because with PCM it is easy to combine digitized voice and digital data into a single, high-speed digital signal and propagate it over either metallic or optical fiber cables. Digital Transmission 279
- 285. Vmax Vmin ts ts ts ts 8-bit word 8-bit word 8-bit word 8-bit word (a) (b) (c) (d) (e) (f) FIGURE 1 Pulse modulation: (a) analog signal; (b) sample pulse; (c) PWM; (d) PPM; (e) PAM; (f) PCM PCM is the only digitally encoded modulation technique shown in Figure 1 that is commonly used for digital transmission. The term pulse code modulation is somewhat of a misnomer, as it is not really a type of modulation but rather a form of digitally coding ana- log signals. With PCM, the pulses are of fixed length and fixed amplitude. PCM is a binary system where a pulse or lack of a pulse within a prescribed time slot represents either a logic 1 or a logic 0 condition. PWM, PPM, and PAM are digital but seldom binary, as a pulse does not represent a single binary digit (bit). Figure 2 shows a simplified block diagram of a single-channel, simplex (one-way only) PCM system. The bandpass filter limits the frequency of the analog input signal to the standard voice-band frequency range of 300 Hz to 3000 Hz. The sample-and-hold cir- Digital Transmission 280
- 286. FIGURE 2 Simplified block diagram of a single-channel, simplex PCM transmission system cuit periodically samples the analog input signal and converts those samples to a multilevel PAM signal. The analog-to-digital converter (ADC) converts the PAM samples to parallel PCM codes, which are converted to serial binary data in the parallel-to-serial converter and then outputted onto the transmission line as serial digital pulses. The transmission line re- peaters are placed at prescribed distances to regenerate the digital pulses. In the receiver, the serial-to-parallel converter converts serial pulses received from the transmission line to parallel PCM codes. The digital-to-analog converter (DAC) con- verts the parallel PCM codes to multilevel PAM signals. The hold circuit is basically a low- pass filter that converts the PAM signals back to its original analog form. Figure 2 also shows several clock signals and sample pulses that will be explained in later sections of this chapter. An integrated circuit that performs the PCM encoding and de- coding functions is called a codec (coder/decoder), which is also described in a later sec- tion of this chapter. 4 PCM SAMPLING The function of a sampling circuit in a PCM transmitter is to periodically sample the con- tinually changing analog input voltage and convert those samples to a series of constant- amplitude pulses that can more easily be converted to binary PCM code. For the ADC to ac- curately convert a voltage to a binary code, the voltage must be relatively constant so that the ADC can complete the conversion before the voltage level changes. If not, theADC would be continually attempting to follow the changes and may never stabilize on any PCM code. Digital Transmission 281
- 287. (a) (b) (c) Input waveform Sample pulse Output waveform FIGURE 3 Natural sampling: (a) input analog signal; (b) sample pulse; (c) sampled output Essentially, there are two basic techniques used to perform the sampling function: natural sampling and flat-top sampling. Natural sampling is shown in Figure 3. Natural sampling is when tops of the sample pulses retain their natural shape during the sample in- terval, making it difficult for an ADC to convert the sample to a PCM code. With natural sampling, the frequency spectrum of the sampled output is different from that of an ideal sample. The amplitude of the frequency components produced from narrow, finite-width sample pulses decreases for the higher harmonics in a (sin x)/x manner. This alters the in- formation frequency spectrum requiring the use of frequency equalizers (compensation fil- ters) before recovery by a low-pass filter. The most common method used for sampling voice signals in PCM systems is flat- top sampling, which is accomplished in a sample-and-hold circuit. The purpose of a sample- and-hold circuit is to periodically sample the continually changing analog input voltage and convert those samples to a series of constant-amplitude PAM voltage levels. With flat-top sampling, the input voltage is sampled with a narrow pulse and then held relatively constant until the next sample is taken. Figure 4 shows flat-top sampling. As the figure shows, the sampling process alters the frequency spectrum and introduces an error called aperture er- ror, which is when the amplitude of the sampled signal changes during the sample pulse time. This prevents the recovery circuit in the PCM receiver from exactly reproducing the original analog signal voltage. The magnitude of error depends on how much the analog signal voltage changes while the sample is being taken and the width (duration) of the sam- ple pulse. Flat-top sampling, however, introduces less aperture distortion than natural sam- pling and can operate with a slower analog-to-digital converter. Figure 5a shows the schematic diagram of a sample-and-hold circuit. The FET acts as a simple analog switch. When turned on, Q1 provides a low-impedance path to deposit the analog sample voltage across capacitor C1. The time that Q1 is on is called the aperture or acquisition time. Essentially, C1 is the hold circuit. When Q1 is off, C1 does not have a com- plete path to discharge through and, therefore, stores the sampled voltage. The storage time of the capacitor is called the A/D conversion time because it is during this time that the ADC converts the sample voltage to a PCM code. The acquisition time should be very short to en- sure that a minimum change occurs in the analog signal while it is being deposited across C1. If the input to the ADC is changing while it is performing the conversion, aperture Digital Transmission 282
- 288. Digital Transmission (a) (b) (c) Input waveform Sample pulse Output waveform FIGURE 4 Flat-top sampling: (a) input analog signal; (b) sample pulse; (c) sampled output FIGURE 5 (a) Sample-and-hold circuit; (b) input and output waveforms 283
- 289. distortion results. Thus, by having a short aperture time and keeping the input to the ADC relatively constant, the sample-and-hold circuit can reduce aperture distortion. Flat-top sam- pling introduces less aperture distortion than natural sampling and requires a slower analog- to-digital converter. Figure 5b shows the input analog signal, the sampling pulse, and the waveform de- veloped across C1. It is important that the output impedance of voltage follower Z1 and the on resistance of Q1 be as small as possible. This ensures that the RC charging time con- stant of the capacitor is kept very short, allowing the capacitor to charge or discharge rap- idly during the short acquisition time. The rapid drop in the capacitor voltage immediately following each sample pulse is due to the redistribution of the charge across C1. The inter- electrode capacitance between the gate and drain of the FET is placed in series with C1 when the FET is off, thus acting as a capacitive voltage-divider network. Also, note the gradual discharge across the capacitor during the conversion time. This is called droop and is caused by the capacitor discharging through its own leakage resistance and the input im- pedance of voltage follower Z2. Therefore, it is important that the input impedance of Z2 and the leakage resistance of C1 be as high as possible. Essentially, voltage followers Z1 and Z2 isolate the sample-and-hold circuit (Q1 and C1) from the input and output circuitry. Example 1 For the sample-and-hold circuit shown in Figure 5a, determine the largest-value capacitor that can be used. Use an output impedance for Z1 of 10 Ω, an on resistance for Q1 of 10 Ω, an acquisition time of 10 μs, a maximum peak-to-peak input voltage of 10 V, a maximum output current from Z1 of 10 mA, and an accuracy of 1%. Solution The expression for the current through a capacitor is Rearranging and solving for C yields where C maximum capacitance (farads) i maximum output current from Z1, 10 mA dv maximum change in voltage across C1, which equals 10 V dt charge time, which equals the aperture time, 10 μs Therefore, The charge time constant for C when Q1 is on is τ RC where τ one charge time constant (seconds) R output impedance of Z1 plus the on resistance of Q1 (ohms) C capacitance value of C1 (farads) Rearranging and solving for C gives us The charge time of capacitor C1 is also dependent on the accuracy desired from the device. The per- cent accuracy and its required RC time constant are summarized as follows: Accuracy (%) Charge Time 10 2.3τ 1 4.6τ 0.1 6.9τ 0.01 9.2τ Cmax τ R Cmax 110 mA2110 μs2 10 V 10 nF C i dt dv i C dv dt Digital Transmission 284
- 290. FIGURE 6 Output spectrum for a sample-and-hold circuit: (a) no aliasing; (b) aliasing distortion For an accuracy of 1%, To satisfy the output current limitations of Z1, a maximum capacitance of 10 nF was required. To sat- isfy the accuracy requirements, 108.7 nF was required. To satisfy both requirements, the smaller- value capacitor must be used. Therefore, C1 can be no larger than 10 nF. 4-1 Sampling Rate The Nyquist sampling theorem establishes the minimum sampling rate (fs) that can be used for a given PCM system. For a sample to be reproduced accurately in a PCM receiver, each cycle of the analog input signal (fa) must be sampled at least twice. Consequently, the min- imum sampling rate is equal to twice the highest audio input frequency. If fs is less than two times fa, an impairment called alias or foldover distortion occurs. Mathematically, the min- imum Nyquist sampling rate is fs ≥ 2fa (1) where fs minimum Nyquist sample rate (hertz) fa maximum analog input frequency (hertz) A sample-and-hold circuit is a nonlinear device (mixer) with two inputs: the sampling pulse and the analog input signal. Consequently, nonlinear mixing (heterodyning) occurs be- tween these two signals. Figure 6a shows the frequency-domain representation of the output spectrum from a sample-and-hold circuit. The output includes the two original inputs (the audio and the fundamental frequency of the sampling pulse), their sum and difference frequencies (fs fa), all the harmonics of fs and fa (2fs, 2fa, 3fs, 3fa, and so on), and their associated cross products (2fs fa, 3fs fa, and so on). Because the sampling pulse is a repetitive waveform, it is made up of a series of har- monically related sine waves. Each of these sine waves is amplitude modulated by the ana- log signal and produces sum and difference frequencies symmetrical around each of the harmonics of fs. Each sum and difference frequency generated is separated from its respec- tive center frequency by fa. As long as fs is at least twice fa, none of the side frequencies from one harmonic will spill into the sidebands of another harmonic, and aliasing does not C 10 μs 4.61202 108.7 nF Digital Transmission 285
- 291. FIGURE 7 Output spectrum for Example 15-2 Table 1 Three-Bit PCM Code Sign Magnitude Decimal Value 1 1 1 3 1 1 0 2 1 0 1 1 1 0 0 0 0 0 0 0 0 0 1 1 0 1 0 2 0 1 1 3 occur. Figure 6b shows the results when an analog input frequency greater than fs/2 mod- ulates fs. The side frequencies from one harmonic fold over into the sideband of another har- monic. The frequency that folds over is an alias of the input signal (hence the names “alias- ing” or “foldover distortion”). If an alias side frequency from the first harmonic folds over into the audio spectrum, it cannot be removed through filtering or any other technique. Example 2 For a PCM system with a maximum audio input frequency of 4 kHz, determine the minimum sample rate and the alias frequency produced if a 5-kHz audio signal were allowed to enter the sample-and- hold circuit. Solution Using Nyquist’s sampling theorem (Equation 1), we have fs ≥ 2fa therefore, fs ≥ 8 kHz If a 5-kHz audio frequency entered the sample-and-hold circuit, the output spectrum shown in Figure 7 is produced. It can be seen that the 5-kHz signal produces an alias frequency of 3 kHz that has been introduced into the original audio spectrum. The input bandpass filter shown in Figure 2 is called an antialiasing or antifoldover filter. Its upper cutoff frequency is chosen such that no frequency greater than one-half the sampling rate is allowed to enter the sample-and-hold circuit, thus eliminating the possi- bility of foldover distortion occurring. With PCM, the analog input signal is sampled, then converted to a serial binary code. The binary code is transmitted to the receiver, where it is converted back to the original ana- log signal. The binary codes used for PCM are n-bit codes, where n may be any positive in- teger greater than 1. The codes currently used for PCM are sign-magnitude codes, where the most significant bit (MSB) is the sign bit and the remaining bits are used for magnitude. Table 1 shows an n-bit PCM code where n equals 3. The most significant bit is used to rep- resent the sign of the sample (logic 1 positive and logic 0 negative). The two re- maining bits represent the magnitude. With two magnitude bits, there are four codes possi- Digital Transmission 286
- 292. Table 2 Three-Bit PCM Code ble for positive numbers and four codes possible for negative numbers. Consequently, there is a total of eight possible codes (23 8). 4-2 Quantization and the Folded Binary Code Quantization is the process of converting an infinite number of possibilities to a finite number of conditions. Analog signals contain an infinite number of amplitude possibili- ties. Thus, converting an analog signal to a PCM code with a limited number of combina- tions requires quantization. In essence, quantization is the process of rounding off the am- plitudes of flat-top samples to a manageable number of levels. For example, a sine wave with a peak amplitude of 5 V varies between 5 V and 5 V passing through every pos- sible amplitude in between. A PCM code could have only eight bits, which equates to only 28 , or 256 combinations. Obviously, to convert samples of a sine wave to PCM requires some rounding off. With quantization, the total voltage range is subdivided into a smaller number of subranges, as shown in Table 2. The PCM code shown in Table 2 is a three-bit sign-magni- tude code with eight possible combinations (four positive and four negative). The leftmost bit is the sign bit (1 and 0 ), and the two rightmost bits represent magnitude. This type of code is called a folded binary code because the codes on the bottom half of the table are a mirror image of the codes on the top half, except for the sign bit. If the negative codes were folded over on top of the positive codes, they would match perfectly. With a folded bi- nary code, each voltage level has one code assigned to it except zero volts, which has two codes, 100 (0) and 000 (0). The magnitude difference between adjacent steps is called the quantization interval or quantum. For the code shown in Table 2, the quantization in- terval is 1 V. Therefore, for this code, the maximum signal magnitude that can be encoded is 3 V (111) or 3 V (011), and the minimum signal magnitude is 1 V (101) or 1 V (001). If the magnitude of the sample exceeds the highest quantization interval, overload distortion (also called peak limiting) occurs. Assigning PCM codes to absolute magnitudes is called quantizing. The magnitude of a quantum is also called the resolution. The resolution is equal to the voltage of the minimum step size, which is equal to the voltage of the least significant bit (Vlsb) of the PCM code. The resolution is the minimum voltage other than 0 V that can be decoded by the digital-to-analog converter in the receiver. The resolution for the PCM code shown in Table 2 is 1 V. The smaller the magnitude of a quantum, the better (smaller) the resolution and the more accurately the quantized signal will resemble the original analog sample. In Table 2, each three-bit code has a range of input voltages that will be converted to that code. For example, any voltage between 0.5 and 1.5 will be converted to the code 101 (1 V). Each code has a quantization range equal to or one-half the magnitude of a quantum except the codes for 0 and 0. The 0-V codes each have an input range equal to only one-half a quantum (0.5 V). Digital Transmission Sign 8 Sub ranges Magnitude Decimal value Quantization range 1 1 1 1 0 0 0 0 1 1 0 0 0 0 1 1 1 0 1 0 0 1 0 1 +3 +2 +1 +0 0 1 2 3 +2.5 V to +3.5 V +1.5 V to +2.5 V +0.5 V to +1.5 V 0 V to +0.5 V 0 V to 0.5 V 0.5 V to 1.5 V 1.5 V to 2.5 V 2.5 V to 3.5 V 287
- 293. FIGURE 8 (a) Analog input signal; (b) sample pulse; (c) PAM signal; (d) PCM code Figure 8 shows an analog input signal, the sampling pulse, the corresponding quan- tized signal (PAM), and the PCM code for each sample. The likelihood of a sample voltage being equal to one of the eight quantization levels is remote. Therefore, as shown in the fig- ure, each sample voltage is rounded off (quantized) to the closest available level and then converted to its corresponding PCM code. The PAM signal in the transmitter is essentially the same PAM signal produced in the receiver. Therefore, any round-off errors in the trans- mitted signal are reproduced when the code is converted back to analog in the receiver. This error is called the quantization error (Qe). The quantization error is equivalent to additive white noise as it alters the signal amplitude. Consequently, quantization error is also called quantization noise (Qn). The maximum magnitude for the quantization error is equal to one- half a quantum (0.5 V for the code shown in Table 2). The first sample shown in Figure 8 occurs at time t1, when the input voltage is exactly 2 V. The PCM code that corresponds to 2 V is 110, and there is no quantization error. Sample 2 occurs at time t2, when the input voltage is 1 V. The corresponding PCM code is 001, and again there is no quantization error. To determine the PCM code for a particu- lar sample voltage, simply divide the voltage by the resolution, convert the quotient to an n-bit binary code, and then add the sign bit. For sample 3 in Figure 9, the voltage at t3 is ap- proximately 2.6 V. The folded PCM code is There is no PCM code for 2.6; therefore, the magnitude of the sample is rounded off to the nearest valid code, which is 111, or 3 V. The rounding-off process results in a quanti- zation error of 0.4 V. sample voltage resolution 2.6 1 2.6 Digital Transmission 288
- 294. FIGURE 9 PAM: (a) input signal; (b) sample pulse; (c) PAM signal The quantized signal shown in Figure 8c at best only roughly resembles the original analog input signal. This is because with a three-bit PCM code, the resolution is rather poor and also because there are only three samples taken of the analog signal. The quality of the PAM signal can be improved by using a PCM code with more bits, reducing the magnitude of a quantum and improving the resolution. The quality can also be improved by sampling the analog signal at a faster rate. Figure 9 shows the same analog input signal shown in Figure 8 except the signal is being sampled at a much higher rate. As the figure shows, the PAM signal resembles the analog input signal rather closely. Figure 10 shows the input-versus-output transfer function for a linear analog-to-dig- ital converter (sometimes called a linear quantizer). As the figure shows for a linear analog input signal (i.e., a ramp), the quantized signal is a staircase function. Thus, as shown in Figure 7c, the maximum quantization error is the same for any magnitude input signal. Example 3 For the PCM coding scheme shown in Figure 8, determine the quantized voltage, quantization error (Qe), and PCM code for the analog sample voltage of 1.07 V. Solution To determine the quantized level, simply divide the sample voltage by resolution and then round the answer off to the nearest quantization level: The quantization error is the difference between the original sample voltage and the quantized level, or Qe 1.07 1 0.07 From Table 2, the PCM code for 1 is 101. 1.07 V 1 V 1.07 1 Digital Transmission 289
- 295. FIGURE 10 Linear input-versus-output transfer curve: (a) linear transfer function; (b) quanti- zation; (c) Qe 4-3 Dynamic Range The number of PCM bits transmitted per sample is determined by several variables, in- cluding maximum allowable input amplitude, resolution, and dynamic range. Dynamic range (DR) is the ratio of the largest possible magnitude to the smallest possible magnitude (other than 0V) that can be decoded by the digital-to-analog converter in the receiver. Math- ematically, dynamic range is (2) where DR dynamic range (unitless ratio) Vmin the quantum value (resolution) Vmax the maximum voltage magnitude that can be discerned by the DACs in the receiver DR Vmax Vmin Digital Transmission 290
- 296. Equation 2 can be rewritten as (3) For the system shown in Table 2, A dynamic range of 3 indicates that the ratio of the largest decoded voltage to the smallest decoded signal voltage is 3 to 1. Dynamic range is generally expressed as a dB value; therefore, (4) For the system shown in Table 2, DR 20 log 3 9.54 dB The number of bits used for a PCM code depends on the dynamic range. The rela- tionship between dynamic range and the number of bits in a PCM code is 2n 1 ≥ DR (5a) and for a minimum number of bits 2n 1 DR (5b) where n number of bits in a PCM code, excluding the sign bit DR absolute value of dynamic range Why 2n 1? One positive and one negative PCM code is used for 0 V, which is not con- sidered for dynamic range. Therefore, 2n DR 1 To solve for the number of bits (n) necessary to produce a dynamic range of 3, convert to logs, log 2n log(DR 1) n log 2 log(DR 1) For a dynamic range of 3, a PCM code with two bits is required. Dynamic range can be ex- pressed in decibels as or DR(dB) 20 log(2n 1) (6) where n is the number of PCM bits. For values of n 4, dynamic range is approximated as DR(dB) ≈ 20 log(2n ) ≈ 20n log(2) ≈ 6n (7) DR1dB2 20 log¢ Vmax Vmin ≤ n log13 12 log 2 0.602 0.301 2 DR 20 log Vmax Vmin DR 3 V 1 V 3 DR Vmax resolution Digital Transmission 291
- 297. Table 3 Dynamic Range versus Number of PCM Magnitude Bits Number of Bits Number of Levels in PCM Code (n) Possible (M 2n ) Dynamic Range (dB) 1 2 6.02 2 4 12 3 8 18.1 4 16 24.1 5 32 30.1 6 64 36.1 7 128 42.1 8 256 48.2 9 512 54.2 10 1024 60.2 11 2048 66.2 12 4096 72.2 13 8192 78.3 14 16,384 84.3 15 32,768 90.3 16 65,536 96.3 Equation 7 indicates that there is approximately 6 dB dynamic range for each magnitude bit in a linear PCM code. Table 3 summarizes dynamic range for PCM codes with n bits for values of n up to 16. Example 4 For a PCM system with the following parameters, determine (a) minimum sample rate, (b) minimum number of bits used in the PCM code, (c) resolution, and (d) quantization error. Maximum analog input frequency 4 kHz Maximum decoded voltage at the receiver 2.55 V Minimum dynamic range 46 dB Solution a. Substituting into Equation 1, the minimum sample rate is fs 2fa 2(4 kHz) 8 kHz b. To determine the absolute value for dynamic range, substitute into Equation 4: 199.5 DR The minimum number of bits is determined by rearranging Equation 5b and solving for n: The closest whole number greater than 7.63 is 8; therefore, eight bits must be used for the magnitude. Because the input amplitude range is 2.55, one additional bit, the sign bit, is required. There- fore, the total number of CM bits is nine, and the total number of PCM codes is 29 512. (There are 255 positive codes, 255 negative codes, and 2 zero codes.) To determine the actual dynamic range, substitute into Equation 6: DR(dB) 20 log(2n 1) 20(log 256 1) 48.13 dB n log1199.5 12 log 2 7.63 102.3 Vmax Vmin 2.3 log Vmax Vmin 46 dB 20 log Vmax Vmin Digital Transmission 292
- 298. c. The resolution is determined by dividing the maximum positive or maximum negative voltage by the number of positive or negative nonzero PCM codes: The maximum quantization error is 4-4 Coding Efficiency Coding efficiency is a numerical indication of how efficiently a PCM code is utilized. Coding efficiency is the ratio of the minimum number of bits required to achieve a cer- tain dynamic range to the actual number of PCM bits used. Mathematically, coding effi- ciency is (8) The coding efficiency for Example 4 is 5 SIGNAL-TO-QUANTIZATION NOISE RATIO The three-bit PCM coding scheme shown in Figures 8 and 9 consists of linear codes, which means that the magnitude change between any two successive codes is the same. Consequently, the magnitude of their quantization error is also the same. The maximum quantization noise is half the resolution (quantum value). Therefore, the worst possible signal voltage-to-quantization noise voltage ratio (SQR) occurs when the input signal is at its minimum amplitude (101 or 001). Mathematically, the worst-case voltage SQR is For the PCM code shown in Figure 8, the worst-case (minimum) SQR occurs for the low- est magnitude quantization voltage (1 V). Therefore, the minimum SQR is or in dB 20 log(2) 6 dB For a maximum amplitude input signal of 3 V (either 111 or 011), the maximum quantiza- tion noise is also equal to the resolution divided by 2. Therefore, the SQR for a maximum input signal is or in dB 20 log 6 15.6 dB SQR1max2 Vmax Qe 3 0.52 6 SQR1min2 1 0.5 2 SQR resolution Qe Vlsb Vlsb2 2 coding efficiency 8.63 9 100 95.89% coding efficiency minimum number of bits 1including sign bit2 actual number of bits 1including sign bit2 100 Qe resolution 2 0.01 V 2 0.005 V resolution Vmax 2n 1 2.55 28 1 2.55 256 1 0.01 V Digital Transmission 293
- 299. From the preceding example, it can be seen that even though the magnitude of the quantization error remains constant throughout the entire PCM code, the percentage error does not; it decreases as the magnitude of the sample increases. The preceding expression for SQR is for voltage and presumes the maximum quan- tization error; therefore, it is of little practical use and is shown only for comparison pur- poses and to illustrate that the SQR is not constant throughout the entire range of sample amplitudes. In reality and as shown in Figure 9, the difference between the PAM waveform and the analog input waveform varies in magnitude. Therefore, the SQR is not constant. Generally, the quantization error or distortion caused by digitizing an analog sample is ex- pressed as an average signal power-to-average noise power ratio. For linear PCM codes (all quantization intervals have equal magnitudes), the signal power-to-quantizing noise power ratio (also called signal-to-distortion ratio or signal-to-noise ratio) is determined by the following formula: (9a) where R resistance (ohms) v rms signal voltage (volts) q quantization interval (volts) v2 /R average signal power (watts) (q2 /12)/R average quantization noise power (watts) If the resistances are assumed to be equal, Equation 8a reduces to 10.8 20 (9b) 6 LINEAR VERSUS NONLINEAR PCM CODES Early PCM systems used linear codes (i.e., the magnitude change between any two suc- cessive steps is uniform). With linear coding, the accuracy (resolution) for the higher- amplitude analog signals is the same as for the lower-amplitude signals, and the SQR for the lower-amplitude signals is less than for the higher-amplitude signals. With voice trans- mission, low-amplitude signals are more likely to occur than large-amplitude signals. Therefore, if there were more codes for the lower amplitudes, it would increase the accu- racy where the accuracy is needed. As a result, there would be fewer codes available for the higher amplitudes, which would increase the quantization error for the larger-amplitude signals (thus decreasing the SQR). Such a coding technique is called nonlinear or nonuniform encoding. With nonlinear encoding, the step size increases with the amplitude of the input signal. Figure 11 shows the step outputs from a linear and a nonlinear analog-to-digital con- verter. Note, with nonlinear encoding, there are more codes at the bottom of the scale than there are at the top, thus increasing the accuracy for the smaller-amplitude signals. Also note that the distance between successive codes is greater for the higher-amplitude signals, thus increasing the quantization error and reducing the SQR. Also, because the ratio of Vmax to Vmin is increased with nonlinear encoding, the dynamic range is larger than with a uniform linear code. It is evident that nonlinear encoding is a compromise; SQR is sacrificed for the higher-amplitude signals to achieve more accuracy for the lower-amplitude signals and to achieve a larger dynamic range. It is difficult to fabricate log v q SQR 10 logc v2 q2 12 d SQR1dB2 10 log v2 R 1q2 122R Digital Transmission 294
- 300. FIGURE 11 (a) Linear versus (b) nonlinear encoding FIGURE 12 Idle channel noise nonlinear analog-to-digital converters; consequently, alternative methods of achieving the same results have been devised. 7 IDLE CHANNEL NOISE During times when there is no analog input signal, the only input to the PAM sampler is random, thermal noise. This noise is called idle channel noise and is converted to a PAM sample just as if it were a signal. Consequently, even input noise is quantized by the ADC. Figure 12 shows a way to reduce idle channel noise by a method called midtread quantiza- tion. With midtread quantizing, the first quantization interval is made larger in amplitude than the rest of the steps. Consequently, input noise can be quite large and still be quantized as a positive or negative zero code. As a result, the noise is suppressed during the encoding process. In the PCM codes described thus far, the lowest-magnitude positive and negative codes have the same voltage range as all the other codes ( or one-half the resolution). This is called midrise quantization. Figure 12 contrasts the idle channel noise transmitted with a midrise PCM code to the idle channel noise transmitted when midtread quantization is used. The advantage of midtread quantization is less idle channel noise. The disadvantage is a larger possible magnitude for Qe in the lowest quantization interval. Digital Transmission 295
- 301. With a folded binary PCM code, residual noise that fluctuates slightly above and be- low 0 V is converted to either a or zero PCM code and, consequently, is eliminated. In systems that do not use the two 0-V assignments, the residual noise could cause the PCM encoder to alternate between the zero code and the minimum or code. Consequently, the decoder would reproduce the encoded noise. With a folded binary code, most of the residual noise is inherently eliminated by the encoder. 8 CODING METHODS There are several coding methods used to quantize PAM signals into 2n levels. These meth- ods are classified according to whether the coding operation proceeds a level at a time, a digit at a time, or a word at a time. 8-1 Level-at-a-Time Coding This type of coding compares the PAM signal to a ramp waveform while a binary counter is being advanced at a uniform rate. When the ramp waveform equals or exceeds the PAM sample, the counter contains the PCM code. This type of coding requires a very fast clock if the number of bits in the PCM code is large. Level-at-a-time coding also requires that 2n sequential decisions be made for each PCM code generated. Therefore, level-at-a-time cod- ing is generally limited to low-speed applications. Nonuniform coding is achieved by using a nonlinear function as the reference ramp. 8-2 Digit-at-a-Time Coding This type of coding determines each digit of the PCM code sequentially. Digit-at-a-time coding is analogous to a balance where known reference weights are used to determine an unknown weight. Digit-at-a-time coders provide a compromise between speed and com- plexity. One common kind of digit-at-a-time coder, called a feedback coder, uses a succes- sive approximation register (SAR). With this type of coder, the entire PCM code word is determined simultaneously. 8-3 Word-at-a-Time Coding Word-at-a-time coders are flash encoders and are more complex; however, they are more suitable for high-speed applications. One common type of word-at-a-time coder uses mul- tiple threshold circuits. Logic circuits sense the highest threshold circuit sensed by the PAM input signal and produce the approximate PCM code. This method is again impractical for large values of n. 9 COMPANDING Companding is the process of compressing and then expanding. With companded systems, the higher-amplitude analog signals are compressed (amplified less than the lower-amplitude signals) prior to transmission and then expanded (amplified more than the lower-amplitude signals) in the receiver. Companding is a means of improving the dynamic range of a com- munications system. Figure 13 illustrates the process of companding. An analog input signal with a dy- namic range of 50 dB is compressed to 25 dB prior to transmission and then, in the receiver, expanded back to its original dynamic range of 50 dB. With PCM, companding may be ac- complished using analog or digital techniques. Early PCM systems used analog compand- ing, whereas more modern systems use digital companding. 9-1 Analog Companding Historically, analog compression was implemented using specially designed diodes in- serted in the analog signal path in a PCM transmitter prior to the sample-and-hold circuit. Digital Transmission 296
- 302. Digital Transmission +20 dB +10 dB –10 dB –20 dB –30 dB Input Transmission media –15 dB –10 dB –5 dB +5 dB +10 dB 25 dB Compressed dynamic range Compression 50 dB Dynamic range 50 dB Dynamic range Expansion 0 dB 0 dB +20 dB +10 dB –10 dB –20 dB –30 dB Output 0 dB FIGURE 13 Basic companding process Analog expansion was also implemented with diodes that were placed just after the low- pass filter in the PCM receiver. Figure 14 shows the basic process of analog companding. In the transmitter, the dy- namic range of the analog signal is compressed, sampled, and then converted to a linear PCM code. In the receiver, the PCM code is converted to a PAM signal, filtered, and then expanded back to its original dynamic range. Different signal distributions require different companding characteristics. For in- stance, voice-quality telephone signals require a relatively constant SQR performance over a wide dynamic range, which means that the distortion must be proportional to signal am- plitude for all input signal levels. This requires a logarithmic compression ratio, which re- quires an infinite dynamic range and an infinite number of PCM codes. Of course, this is impossible to achieve. However, there are two methods of analog companding currently be- ing used that closely approximate a logarithmic function and are often called log-PCM codes. The two methods are μ-law and the A-law companding. 9-1-1 μ-Law companding. In the United States and Japan, μ-law companding is used. The compression characteristics for μ-law is (10) Vout Vmax ln11 μVinVmax 2 ln11 μ2 297
- 303. FIGURE 15 μ-law compression char- acteristics where Vmax maximum uncompressed analog input amplitude (volts) Vin amplitude of the input signal at a particular instant of time (volts) μ parameter used to define the amount of compression (unitless) Vout compressed output amplitude (volts) Figure 15 shows the compression curves for several values of μ. Note that the higher the μ, the more compression. Also note that for μ 0, the curve is linear (no compression). The parameter μ determines the range of signal power in which the SQR is relatively constant. Voice transmission requires a minimum dynamic range of 40 dB and a seven-bit PCM code. For a relatively constant SQR and a 40-dB dynamic range, a μ 100 is required. The early Bell System PCM systems used a seven-bit code with a μ 100. However, the most recent PCM systems use an eight-bit code and a μ 255. Digital Transmission FIGURE 14 PCM system with analog companding 298
- 304. Example 5 For a compressor with a μ 255, determine a. The voltage gain for the following relative values of Vin: Vmax, 0.75 Vmax, 0.5 Vmax, and 0.25 Vmax. b. The compressed output voltage for a maximum input voltage of 4 V. c. Input and output dynamic ranges and compression. Solution a. Substituting into Equation 10, the following voltage gains are achieved for the given input magnitudes: Compressed Vin Voltage Gain Vmax 1.00 0.75 Vmax 1.26 0.5 Vmax 1.75 0.25 Vmax 3.00 b. Using the compressed voltage gains determined in step (a), the output voltage is simply the input voltage times the compression gain: Vin Voltage Gain Vout Vmax 4 V 1.00 4.00 V 0.75 Vmax 3 V 1.26 3.78 V 0.50 Vmax 2 V 1.75 3.50 V 0.25 Vmax 1 V 3.00 3 00 V c. Dynamic range is calculated by substituting into Equation 4: compression input dynamic range minus output dynamic range 12 dB 2.5 dB 9.5 dB To restore the signals to their original proportions in the receiver, the compressed voltages are expanded by passing them through an amplifier with gain characteristics that are the complement of those in the compressor. For the values given in Example 5, the volt- age gains in the receiver are as follows: Expanded Vin Voltage Gain Vmax 1.00 0.75 Vmax 0.79 0.5 Vmax 0.57 0.25 Vmax 0.33 The overall circuit gain is simply the product of the compression and expansion fac- tors, which equals one for all input voltage levels. For the values given in Example 5, Vin Vmax 1 1 1 Vin 0.75 Vmax 1.26 0.79 ⬵ 1 Vin 0.5 Vmax 1.75 0.57 ⬵ 1 Vin 0.25 Vmax 3 0.33 ⬵ 1 9-1-2 A-law companding. In Europe, the ITU-T has established A-law compand- ing to be used to approximate true logarithmic companding. For an intended dynamic output dynamic range 20 log 4 3 2.5 dB input dynamic range 20 log 4 1 12 dB Digital Transmission 299
- 305. FIGURE 16 Digitally companded PCM system range, A-law companding has a slightly flatter SQR than μ-law. A-law companding, how- ever, is inferior to μ-law in terms of small-signal quality (idle channel noise). The com- pression characteristic for A-law companding is (11a) (11b) 9-2 Digital Companding Digital companding involves compression in the transmitter after the input sample has been converted to a linear PCM code and then expansion in the receiver prior to PCM decoding. Figure 16 shows the block diagram for a digitally companded PCM system. Withdigitalcompanding,theanalogsignalisfirstsampledandconvertedtoalinearPCM codeandthenthelinearcodeisdigitallycompressed.Inthereceiver,thecompressedPCMcode is expanded and then decoded (i.e., converted back to analog). The most recent digitally com- pressed PCM systems use a 12-bit linear PCM code and an eight-bit compressed PCM code. The compression and expansion curves closely resemble the analog μ-law curves with a μ 255 by approximating the curve with a set of eight straight-line segments (segments 0 through7).Theslopeofeachsuccessivesegmentisexactlyone-halfthatoftheprevioussegment. Figure 17 shows the 12-bit-to-8-bit digital compression curve for positive values only. The curve for negative values is identical except the inverse. Although there are 16 segments (eight positive and eight negative), this scheme is often called 13-segment com- pression because the curve for segments 0, 1, 0, and 1 is a straight line with a con- stant slope and is considered as one segment. The digital companding algorithm for a 12-bit linear-to-8-bit compressed code is ac- tually quite simple. The eight-bit compressed code consists of a sign bit, a three-bit segment identifier, and a 10-bit magnitude code that specifies the quantization interval within the specified segment (see Figure 18a). In the μ255-encoding table shown in Figure 18b, the bit positions designated with an X are truncated during compression and subsequently lost. Bits designated A, B, C, and 1 1n1AVinVmax 2 1 ln A 1 A Vin Vmax 1 Vout Vmax AVinVmax 1 ln A 0 Vin Vmax 1 A Digital Transmission 300
- 306. Digital Transmission FIGURE 17 μ255 compression characteristics (positive values only) FIGURE 18 12-bit-to-8-bit digital companding: (a) 8-bit μ255 compressed code format; (b) μ255 encoding table; (c) μ255 decoding table 301
- 307. D are transmitted as is. The sign bit is also transmitted as is. Note that for segments 0 and 1, the encoded 12-bit PCM code is duplicated exactly at the output of the decoder (compare Figures 18b and c), whereas for segment 7, only the most significant six bits are duplicated. With 11 magnitude bits, there are 2048 possible codes, but they are not equally distributed among the eight segments. There are 16 codes in segment 0 and 16 codes in segment 1. In each subsequent segment, the number of codes doubles (i.e., segment 2 has 32 codes; seg- ment 3 has 64 codes, and so on). However, in each of the eight segments, only 16 12-bit codes can be produced. Consequently, in segments 0 and 1, there is no compression (of the 16 possible codes, all 16 can be decoded). In segment 2, there is a compression ratio of 2:1 (of the 32 possible codes, only 16 can be decoded). In segment 3, there is a 4:1 compres- sion ratio (64 codes to 16 codes). The compression ratio doubles with each successive seg- ment. The compression ratio in segment 7 is 1024/16, or 64:1. The compression process is as follows. The analog signal is sampled and converted to a linear 12-bit sign-magnitude code. The sign bit is transferred directly to an eight-bit compressed code. The segment number in the eight-bit code is determined by counting the number of leading 0s in the 11-bit magnitude portion of the linear code beginning with the most significant bit. Subtract the number of leading 0s (not to exceed 7) from 7. The result is the segment number, which is converted to a three-bit binary number and inserted into the eight-bit compressed code as the segment identifier. The four magnitude bits (A, B, C, and D) represent the quantization interval (i.e., subsegments) and are substituted into the least significant four bits of the 8-bit compressed code. Essentially, segments 2 through 7 are subdivided into smaller subsegments. Each seg- ment consists of 16 subsegments, which correspond to the 16 conditions possible for bits A, B, C, and D (0000 to 1111). In segment 2, there are two codes per subsegment. In seg- ment 3, there are four. The number of codes per subsegment doubles with each subsequent segment. Consequently, in segment 7, each subsegment has 64 codes. Figure 19 shows the breakdown of segments versus subsegments for segments 2, 5, and 7. Note that in each subsegment, all 12-bit codes, once compressed and expanded, yield a single 12-bit code. In the decoder, the most significant of the truncated bits is reinserted as a logic 1. The remaining truncated bits are reinserted as 0s. This ensures that the maxi- mum magnitude of error introduced by the compression and expansion process is mini- mized. Essentially, the decoder guesses what the truncated bits were prior to encoding. The most logical guess is halfway between the minimum- and maximum-magnitude codes. For example, in segment 6, the five least significant bits are truncated during compression; therefore, in the receiver, the decoder must try to determine what those bits were. The pos- sibilities include any code between 00000 and 11111. The logical guess is 10000, approx- imately half the maximum magnitude. Consequently, the maximum compression error is slightly more than one-half the maximum magnitude for that segment. Example 6 Determine the 12-bit linear code, the eight-bit compressed code, the decoded 12-bit code, the quan- tization error, and the compression error for a resolution of 0.01 V and analog sample voltages of (a) 0.053 V, (b) 0.318 V, and (c) 10.234 V Solution a. To determine the 12-bit linear code, simply divide the sample voltage by the resolu- tion, round off the quotient, and then convert the result to a 12-bit sign-magnitude code: which is rounded off to 5 producing a quantization error Qe 0.3(0.01 V) 0.003 V A B C D 12-bit linear code 1 0 0 0 0 0 0 0 0 1 0 1 sign bit 11-bit magnitude bits 00000000101 5 (1 ) 0.053 V 0.01 V 5.3, Digital Transmission —› ‹ —————–——–———— › 302
- 308. To determine the 8-bit 1 0 0 0 0 0 0 0 0 1 0 1 compressed code, 1 (7 7 0) A B C D sign unit quantization bit identifier interval () (segment 0) (5) 8-bit compressed code 1 0 0 0 0 1 0 1 To determine the 12-bit 1 0 0 0 0 1 0 1 recovered code, simply s (000 segment 0) A B C D reverse the process: sign segment 0 quantization bit has interval () seven leading 0s (0101 5) 12-bit recovered code 1 0 0 0 0 0 0 0 0 1 0 1 5 recovered voltage 5(0.01) 0.05 Digital Transmission FIGURE 19 12-bit segments divided into subsegments: (a) segment 7; (Continued) 303
- 309. As Example 6 shows, the recovered 12-bit code (5) is exactly the same as the original 12-bit linear code (5). Therefore, the decoded voltage (0.05 V) is the same as the original encoded voltage (0.5). This is true for all codes in segments 0 and 1. Thus, there is no compression error in segments 0 and 1, and the only error produced is from the quantizing process (for this example, the quantiza- tion error Qe 0.003 V). b. To determine the 12-bit linear code, which is rounded off to 32, producing a quantization error Qe 0.2 (0.01 V) 0.002 V A B C D 12-bit linear code 0 0 0 0 0 0 1 0 0 0 0 0 11-bit magnitude bits sign bit (0 ) 0.318 V 0.01 V 31.8, Digital Transmission —› ——› ——› FIGURE 19 (Continued) (b) segment 5 304
- 310. To determine the 8-bit 0 0 0 0 0 0 1 0 0 0 0 0 compressed code, 0 (7 5 2) A B C D X sign unit quantization truncated bit identifier interval () (segment 2) (0) eight-bit compressed code 0 0 1 0 0 0 0 0 Again, to determine 0 0 1 0 0 0 0 0 the 12-bit recovered (7 2 5) A B C D code, simply reverse sign segment 5 quantization the process: bit has five interval () leading 0s (0000 0) A B C D 12-bit recovered code 0 0 0 0 0 0 1 0 0 0 0 1 33 s inserted inserted decoded voltage 33(0.1) 0.33 V Digital Transmission —› —› —› FIGURE 19 (Continued) (c) segment 2 305
- 311. Note the two inserted ones in the recovered 12-bit code. The least significant bit is determined from the decoding table shown in Figure 18c. As the figure shows, in the receiver the most significant of the truncated bits is always set (1), and all other truncated bits are cleared (0s). For segment 2 codes, there is only one truncated bit; thus, it is set in the receiver. The inserted 1 in bit position 6 was dropped during the 12-bit-to-8-bit conversion process, as transmission of this bit is redundant because if it were not a 1, the sample would not be in that segment. Consequently, for all segments except segments 0 and 1, a 1 is automatically inserted between the reinserted 0s and the ABCD bits. For this example, there are two errors: the quantization error and the compression error. The quantization error is due to rounding off the sample voltage in the encoder to the closest PCM code, and the compression error is caused by forcing the truncated bit to be a 1 in the receiver. Keep in mind that the two errors are not always additive, as they could cause errors in the oppo- site direction and actually cancel each other. The worst-case scenario would be when the two errors were in the same direction and at their maximum values. For this example, the combined error was 0.33 V 0.318 V 0.012 V. The worst possible error in segments 0 and 1 is the maximum quanti- zation error, or half the magnitude of the resolution. In segments 2 through 7, the worst possible er- ror is the sum of the maximum quantization error plus the magnitude of the most significant of the truncated bits. c. To determine the 12-bit linear code, which is rounded off to 1023, producing a quantization error Qe 0.4(0.01 V) 0.004 V A B C D 12-bit linear code 1 0 1 1 1 1 1 1 1 1 1 1 11-bit magnitude bits sign bit (1 ) To determine the 8-bit 1 0 1 1 1 1 1 1 1 1 1 1 compressed code, 1 A B C D X X X X X truncated 8-bit compressed code 1 1 1 0 1 1 1 1 To determine the 12-bit 1 1 1 0 1 1 1 1 recovered code, simply s segment 6 A B C D 12-bit recovered code 1 0 1 1 1 1 1 1 0 0 0 0 1008 A B C D s inserted inserted decoded voltage 1008(0.01) 10.08 V The difference between the original 12-bit code and the decoded 12-bit code is 10.23 10.08 0.15 or For this example, there are again two errors: a quantization error of 0.004 V and a compression error of 0.15 V. The combined error is 10.234 V 10.08 V 0.154 V. 9-3 Digital Compression Error As seen in Example 6, the magnitude of the compression error is not the same for all sam- ples. However, the maximum percentage error is the same in each segment (other than seg- ments 0 and 1, where there is no compression error). For comparison purposes, the follow- ing formula is used for computing the percentage error introduced by digital compression: (12) % error 12-bit encoded voltage 12-bit decoded voltage 12-bit decoded voltage 100 1011 1111 1111 1011 1111 0000 1111 1510.012 0.15 V 10.234 V 0.01 V 1023.4, Digital Transmission —› ———› —————— › —› —› 306
- 312. Example 7 The maximum percentage error will occur for the smallest number in the lowest subsegment within any given segment. Because there is no compression error in segments 0 and 1, for segment 3 the max- imum percentage error is computed as follows: transmit 12-bit code s 0 0 0 0 1 0 0 0 0 0 0 receive 12-bit code s 0 0 0 0 1 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 1 0 and for segment 7 transmit 12-bit code s 1 0 0 0 0 0 0 0 0 0 0 receive 12-bit code s 1 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 As Example 7 shows, the maximum magnitude of error is higher for segment 7; how- ever, the maximum percentage error is the same for segments 2 through 7. Consequently, the maximum SQR degradation is the same for each segment. Although there are several ways in which the 12-bit-to-8-bit compression and 8-bit- to-12-bit expansion can be accomplished with hardware, the simplest and most economical method is with a lookup table in ROM (read-only memory). Essentially every function performed by a PCM encoder and decoder is now accom- plished with a single integrated-circuit chip called a codec. Most of the more recently de- veloped codecs are called combo chips, as they include an antialiasing (bandpass) filter, a sample-and-hold circuit, and an analog-to-digital converter in the transmit section and a digital-to-analog converter, a hold circuit, and a bandpass filter in the receive section. 10 VOCODERS The PCM coding and decoding processes described in the preceding sections were con- cerned primarily with reproducing waveforms as accurately as possible. The precise nature of the waveform was unimportant as long as it occupied the voice-band frequency range. When digitizing speech signals only, special voice encoders/decoders called vocoders are often used. To achieve acceptable speech communications, the short-term power spectrum of the speech information is all that must be preserved. The human ear is relatively insen- sitive to the phase relationship between individual frequency components within a voice waveform. Therefore, vocoders are designed to reproduce only the short-term power spec- trum, and the decoded time waveforms often only vaguely resemble the original input sig- nal.Vocoders cannot be used in applications where analog signals other than voice are pres- ent, such as output signals from voice-band data modems. Vocoders typically produce unnatural sounding speech and, therefore, are generally used for recorded information, such as “wrong number” messages, encrypted voice for transmission over analog telephone circuits, computer output signals, and educational games. 冟1024 1056冟 1056 100 3.03% % error 冟10000000000 10000100000冟 10000100000 100 冟64 66冟 66 100 3.03% % error 冟1000000 1000010冟 1000010 100 Digital Transmission 307
- 313. Digital Transmission The purpose of a vocoder is to encode the minimum amount of speech information necessary to reproduce a perceptible message with fewer bits than those needed by a con- ventional encoder/decoder. Vocoders are used primarily in limited bandwidth applications. Essentially, there are three vocoding techniques available: the channel vocoder, the formant vocoder, and the linear predictive coder. 10-1 Channel Vocoders The first channel vocoder was developed by Homer Dudley in 1928. Dudley’s vocoder compressed conventional speech waveforms into an analog signal with a total bandwidth of approximately 300 Hz. Present-day digital vocoders operate at less than 2 kbps. Digital channel vocoders use bandpass filters to separate the speech waveform into narrower sub- bands. Each subband is full-wave rectified, filtered, and then digitally encoded. The en- coded signal is transmitted to the destination receiver, where it is decoded. Generally speak- ing, the quality of the signal at the output of a vocoder is quite poor. However, some of the more advanced channel vocoders operate at 2400 bps and can produce a highly intelligible, although slightly synthetic sounding speech. 10-2 Formant Vocoders A formant vocoder takes advantage of the fact that the short-term spectral density of typi- cal speech signals seldom distributes uniformly across the entire voice-band spectrum (300 Hz to 3000 Hz). Instead, the spectral power of most speech energy concentrates at three or four peak frequencies called formants. A formant vocoder simply determines the location of these peaks and encodes and transmits only the information with the most sig- nificant short-term components. Therefore, formant vocoders can operate at lower bit rates and, thus, require narrower bandwidths. Formant vocoders sometimes have trouble track- ing changes in the formants. However, once the formants have been identified, a formant vocoder can transfer intelligible speech at less than 1000 bps. 10-3 Linear Predictive Coders A linear predictive coder extracts the most significant portions of speech information directly from the time waveform rather than from the frequency spectrum as with the channel and formant vocoders. A linear predictive coder produces a time-varying model of the vocal tract excitation and transfer function directly from the speech wave- form. At the receive end, a synthesizer reproduces the speech by passing the specified excitation through a mathematical model of the vocal tract. Linear predictive coders provide more natural sounding speech than either the channel or the formant vocoder. Linear predictive coders typically encode and transmit speech at between 1.2 kbps and 2.4 kbps. 11 PCM LINE SPEED Line speed is simply the data rate at which serial PCM bits are clocked out of the PCM en- coder onto the transmission line. Line speed is dependent on the sample rate and the num- ber of bits in the compressed PCM code. Mathematically, line speed is (13) where line speed the transmission rate in bits per second samples/second sample rate (fs) bits/sample number of bits in the compressed PCM code line speed samples second bits sample 308
- 314. Digital Transmission Example 8 For a single-channel PCM system with a sample rate fs 6000 samples per second and a seven-bit compressed PCM code, determine the line speed: Solution 12 DELTA MODULATION PCM Delta modulation uses a single-bit PCM code to achieve digital transmission of analog sig- nals. With conventional PCM, each code is a binary representation of both the sign and the magnitude of a particular sample. Therefore, multiple-bit codes are required to represent the many values that the sample can be. With delta modulation, rather than transmit a coded rep- resentationofthesample,onlyasinglebitistransmitted,whichsimplyindicateswhetherthat sample is larger or smaller than the previous sample. The algorithm for a delta modulation system is quite simple. If the current sample is smaller than the previous sample, a logic 0 is transmitted. If the current sample is larger than the previous sample, a logic 1 is transmitted. 12-1 Delta Modulation Transmitter Figure 20 shows a block diagram of a delta modulation transmitter. The input analog is sam- pled and converted to a PAM signal, which is compared with the output of the DAC. The output of the DAC is a voltage equal to the regenerated magnitude of the previous sample, which was stored in the up–down counter as a binary number. The up–down counter is in- cremented or decremented depending on whether the previous sample is larger or smaller than the current sample. The up–down counter is clocked at a rate equal to the sample rate. Therefore, the up–down counter is updated after each comparison. Figure 21 shows the ideal operation of a delta modulation encoder. Initially, the up–down counter is zeroed, and the DAC is outputting 0 V. The first sample is taken, con- verted to a PAM signal, and compared with zero volts. The output of the comparator is a 42,000 bps line speed 6000 samples second 7 bits sample FIGURE 20 Delta modulation transmitter 309
- 315. Digital Transmission FIGURE 21 Ideal operation of a delta modulation encoder logic 1 condition (V), indicating that the current sample is larger in amplitude than the previous sample. On the next clock pulse, the up–down counter is incremented to a count of 1. The DAC now outputs a voltage equal to the magnitude of the minimum step size (res- olution). The steps change value at a rate equal to the clock frequency (sample rate). Con- sequently, with the input signal shown, the up–down counter follows the input analog sig- nal up until the output of the DAC exceeds the analog sample; then the up–down counter will begin counting down until the output of the DAC drops below the sample amplitude. In the idealized situation (shown in Figure 21), the DAC output follows the input signal. Each time the up–down counter is incremented, a logic 1 is transmitted, and each time the up–down counter is decremented, a logic 0 is transmitted. 12-2 Delta Modulation Receiver Figure 22 shows the block diagram of a delta modulation receiver.As you can see, the receiver is almost identical to the transmitter except for the comparator. As the logic 1s and 0s are re- ceived, the up–down counter is incremented or decremented accordingly. Consequently, the output of the DAC in the decoder is identical to the output of the DAC in the transmitter. With delta modulation, each sample requires the transmission of only one bit; there- fore, the bit rates associated with delta modulation are lower than conventional PCM sys- tems. However, there are two problems associated with delta modulation that do not occur with conventional PCM: slope overload and granular noise. 12-2-1 Slope overload. Figure 23 shows what happens when the analog input sig- nal changes at a faster rate than the DAC can maintain. The slope of the analog signal is greater than the delta modulator can maintain and is called slope overload. Increasing the clock frequency reduces the probability of slope overload occurring. Another way to pre- vent slope overload is to increase the magnitude of the minimum step size. FIGURE 22 Delta modulation receiver 310
- 316. Digital Transmission FIGURE 23 Slope overload distortion FIGURE 24 Granular noise 12-2-2 Granular noise. Figure 24 contrasts the original and reconstructed signals associated with a delta modulation system. It can be seen that when the original analog in- put signal has a relatively constant amplitude, the reconstructed signal has variations that were not present in the original signal. This is called granular noise. Granular noise in delta modulation is analogous to quantization noise in conventional PCM. Granular noise can be reduced by decreasing the step size. Therefore, to reduce the granular noise, a small resolution is needed, and to reduce the possibility of slope overload occurring, a large resolution is required. Obviously, a compromise is necessary. Granular noise is more prevalent in analog signals that have gradual slopes and whose amplitudes vary only a small amount. Slope overload is more prevalent in analog signals that have steep slopes or whose amplitudes vary rapidly. 13 ADAPTIVE DELTA MODULATION PCM Adaptive delta modulation is a delta modulation system where the step size of the DAC is automatically varied, depending on the amplitude characteristics of the analog input signal. Figure 25 shows how an adaptive delta modulator works. When the output of the trans- mitter is a string of consecutive 1s or 0s, this indicates that the slope of the DAC output is FIGURE 25 Adaptive delta modulation 311
- 317. Digital Transmission less than the slope of the analog signal in either the positive or the negative direction. Essentially, the DAC has lost track of exactly where the analog samples are, and the possibility of slope overload occurring is high. With an adaptive delta modulator, after a predetermined number of consecutive 1s or 0s, the step size is automatically in- creased. After the next sample, if the DAC output amplitude is still below the sample amplitude, the next step is increased even further until eventually the DAC catches up with the analog signal. When an alternative sequence of 1s and 0s is occurring, this in- dicates that the possibility of granular noise occurring is high. Consequently, the DAC will automatically revert to its minimum step size and, thus, reduce the magnitude of the noise error. A common algorithm for an adaptive delta modulator is when three consecutive 1s or 0s occur, the step size of the DAC is increased or decreased by a factor of 1.5. Various other algorithms may be used for adaptive delta modulators, depending on particular system re- quirements. 14 DIFFERENTIAL PCM In a typical PCM-encoded speech waveform, there are often successive samples taken in which there is little difference between the amplitudes of the two samples. This necessitates transmitting several identical PCM codes, which is redundant. Differential pulse code mod- ulation (DPCM) is designed specifically to take advantage of the sample-to-sample redun- dancies in typical speech waveforms. With DPCM, the difference in the amplitude of two successive samples is transmitted rather than the actual sample. Because the range of sam- ple differences is typically less than the range of individual samples, fewer bits are required for DPCM than conventional PCM. Figure 26 shows a simplified block diagram of a DPCM transmitter. The analog input signal is bandlimited to one-half the sample rate, then compared with the preceding accu- mulated signal level in the differentiator. The output of the differentiation is the difference between the two signals. The difference is PCM encoded and transmitted. The ADC oper- ates the same as in a conventional PCM system, except that it typically uses fewer bits per sample. Figure 27 shows a simplified block diagram of a DPCM receiver. Each received sam- ple is converted back to analog, stored, and then summed with the next sample received. In the receiver shown in Figure 27, the integration is performed on the analog signals, although it could also be performed digitally. FIGURE 26 DPCM transmitter 312
- 318. Digital Transmission Serial DPCM in Parallel DPCM Sum signal out PAM Serial to parallel converter PAM PAM Analog out Adder (integrator) + Hold circuit Digital to analog converter Low-pass filter + FIGURE 27 DPCM receiver 15 PULSE TRANSMISSION All digital carrier systems involve the transmission of pulses through a medium with a fi- nite bandwidth. A highly selective system would require a large number of filter sections, which is impractical. Therefore, practical digital systems generally utilize filters with bandwidths that are approximately 30% or more in excess of the ideal Nyquist bandwidth. Figure 28a shows the typical output waveform from a bandlimited communications chan- nel when a narrow pulse is applied to its input. The figure shows that bandlimiting a pulse causes the energy from the pulse to be spread over a significantly longer time in the form of secondary lobes. The secondary lobes are called ringing tails. The output frequency spectrum corresponding to a rectangular pulse is referred to as a (sin x)/x response and is given as (14) where ω 2πf (radians) T pulse width (seconds) Figure 28b shows the distribution of the total spectrum power. It can be seen that ap- proximately 90% of the signal power is contained within the first spectral null (i.e., f 1/T). Therefore, the signal can be confined to a bandwidth B 1/T and still pass most of the energy from the original waveform. In theory, only the amplitude at the middle of each pulse interval needs to be preserved. Therefore, if the bandwidth is confined to B 1/2T, the maximum signaling rate achievable through a low-pass filter with a specified bandwidth without causing excessive distortion is given as the Nyquist rate and is equal to twice the bandwidth. Mathematically, the Nyquist rate is R 2B (15) where R signaling rate 1/T B specified bandwidth f 12 1T2 sin1T22 T2 313
- 319. Digital Transmission FIGURE 28 Pulse response: (a) typical pulse response of a bandlimited filter; (b) spectrum of square pulse with duration 1/T 15-1 Intersymbol Interference Figure 29 shows the input signal to an ideal minimum bandwidth, low-pass filter. The in- put signal is a random, binary nonreturn-to-zero (NRZ) sequence. Figure 29b shows the output of a low-pass filter that does not introduce any phase or amplitude distortion. Note that the output signal reaches its full value for each transmitted pulse at precisely the cen- ter of each sampling interval. However, if the low-pass filter is imperfect (which in reality it will be), the output response will more closely resemble that shown in Figure 29c. At the sampling instants (i.e., the center of the pulses), the signal does not always attain the max- imum value. The ringing tails of several pulses have overlapped, thus interfering with the major pulse lobe. Assuming no time delays through the system, energy in the form of spu- rious responses from the third and fourth impulses from one pulse appears during the sam- pling instant (T 0) of another pulse. This interference is commonly called intersymbol in- terference, or simply ISI. ISI is an important consideration in the transmission of pulses over circuits with a limited bandwidth and a nonlinear phase response. Simply stated, rec- tangular pulses will not remain rectangular in less than an infinite bandwidth. The nar- rower the bandwidth, the more rounded the pulses. If the phase distortion is excessive, the pulse will tilt and, consequently, affect the next pulse. When pulses from more than 314
- 320. Digital Transmission FIGURE 29 Pulse response: (a) NRZ input signal; (b) output from a perfect filter; (c) output from an imperfect filter one source are multiplexed together, the amplitude, frequency, and phase responses become even more critical. ISI causes crosstalk between channels that occupy adjacent time slots in a time-division-multiplexed carrier system. Special filters called equalizers are inserted in the transmission path to “equalize” the distortion for all frequencies, creating a uniform transmission medium and reducing transmission impairments. The four primary causes of ISI are as follows: 1. Timing inaccuracies. In digital transmission systems, transmitter timing inaccura- cies cause intersymbol interference if the rate of transmission does not conform to the ringing frequency designed into the communications channel. Generally, timing inaccura- cies of this type are insignificant. Because receiver clocking information is derived from the received signals, which are contaminated with noise, inaccurate sample timing is more likely to occur in receivers than in transmitters. 2. Insufficient bandwidth. Timing errors are less likely to occur if the transmission rate is well below the channel bandwidth (i.e., the Nyquist bandwidth is significantly be- low the channel bandwidth).As the bandwidth of a communications channel is reduced, the ringing frequency is reduced, and intersymbol interference is more likely to occur. 315
- 321. Digital Transmission 3. Amplitude distortion. Filters are placed in a communications channel to bandlimit signals and reduce or eliminate predicted noise and interference. Filters are also used to pro- duce a specific pulse response. However, the frequency response of a channel cannot always be predicted absolutely. When the frequency characteristics of a communications channel depart from the normal or expected values, pulse distortion results. Pulse distortion occurs when the peaks of pulses are reduced, causing improper ringing frequencies in the time do- main. Compensation for such impairments is called amplitude equalization. 4. Phase distortion. A pulse is simply the superposition of a series of harmonically related sine waves with specific amplitude and phase relationships. Therefore, if the rela- tive phase relations of the individual sine waves are altered, phase distortion occurs. Phase distortion occurs when frequency components undergo different amounts of time delay while propagating through the transmission medium. Special delay equalizers are placed in the transmission path to compensate for the varying delays, thus reducing the phase distor- tion. Phase equalizers can be manually adjusted or designed to automatically adjust them- selves to varying transmission characteristics. 15-2 Eye Patterns The performance of a digital transmission system depends, in part, on the ability of a re- peater to regenerate the original pulses. Similarly, the quality of the regeneration process depends on the decision circuit within the repeater and the quality of the signal at the input to the decision circuit. Therefore, the performance of a digital transmission system can be measured by displaying the received signal on an oscilloscope and triggering the time base at the data rate. Thus, all waveform combinations are superimposed over adjacent signal- ing intervals. Such a display is called an eye pattern or eye diagram.An eye pattern is a con- venient technique for determining the effects of the degradations introduced into the pulses as they travel to the regenerator. The test setup to display an eye pattern is shown in Figure 30. The received pulse stream is fed to the vertical input of the oscilloscope, and the sym- bol clock is fed to the external trigger input, while the sweep rate is set approximately equal to the symbol rate. Figure 31 shows an eye pattern generated by a symmetrical waveform for ternary sig- nals in which the individual pulses at the input to the regenerator have a cosine-squared shape. In an m-level system, there will be m 1 separate eyes. The vertical lines labeled 1, 0, and 1 correspond to the ideal received amplitudes. The horizontal lines, separated by the signaling interval, T, correspond to the ideal decision times. The decision levels for the regenerator are represented by crosshairs. The vertical hairs represent the decision time, whereas the horizontal hairs represent the decision level. The eye pattern shows the quality of shaping and timing and discloses any noise and errors that might be present in the line equalization. The eye opening (the area in the middle of the eye pattern) defines a boundary within which no waveform trajectories can exist under any code-pattern condition. The eye FIGURE 30 Eye diagram mea- surement setup 316
- 322. Digital Transmission FIGURE 31 Eye diagram opening is a function of the number of code levels and the intersymbol interference caused by the ringing tails of any preceding or succeeding pulses. To regenerate the pulse sequence without error, the eye must be open (i.e., a decision area must exist), and the decision crosshairs must be within the open area. The effect of pulse degradation is a reduction in the size of the ideal eye. In Figure 31, it can be seen that at the center of the eye (i.e., the sam- pling instant) the opening is about 90%, indicating only minor ISI degradation due to filter- ing imperfections.The small degradation is due to the nonideal Nyquist amplitude and phase characteristics of the transmission system. Mathematically, the ISI degradation is (16) where H ideal vertical opening (cm) h degraded vertical opening (cm) For the eye diagram shown in Figure 31, In Figure 31,it can also be seen that the overlapping signal pattern does not cross the hor- izontal zero line at exact integer multiples of the symbol clock. This is an impairment known as data transition jitter. This jitter has an effect on the symbol timing (clock) recovery circuit and,ifexcessive,maysignificantlydegradetheperformanceofcascadedregenerativesections. 16 SIGNAL POWER IN BINARY DIGITAL SIGNALS Because binary digital signals can originate from literally scores of different types of data sources, it is impossible to predict which patterns or sequences of bits are most likely to oc- cur over a given period of time in a given system. Thus, for signal analysis purposes, it is gen- erally assumed that there is an equal probability of the occurrence of a 1 and a 0. Therefore, 20 log 90 100 0.915 dB 1ISI degradation2 ISI 20 log h H 317
- 323. Digital Transmission power can be averaged over an entire message duration, and the signal can be modeled as a continuous sequence of alternating 1s and 0s as shown in Figure 32. Figure 32a shows a stream of rectangularly shaped pulses with a pulse width-to-pulse duration ratio τ/T less than 0.5, and Figure 32b shows a stream of square wave pulses with a τ/T ratio of 0.5. The normalized (R 1) average power is derived for signal f(t) from (17) where T is the period of integration. If f(t) is a periodic signal with period T0, then Equation 17 reduces to (18) If rectangular pulses of amplitude V with a τ/T ratio of 0.5 begin at t 0, then (19) Thus, from Equation 18, (20) and Because the effective rms value of a periodic wave is found from P (Vrms)2 /R, the rms voltage for a rectangular pulse is (21) Because With the square wave shown in Figure 32, τ/T 0.5, therefore, Thus, the rms voltage for the square wave is the same as for sine waves, Vrms V12. P V 2 2R. P 1Vrms 22 R, P 12τTV22 R 1τV2 21TR2. Vrms B τ T 1V2 P ¢ τ T ≤ V2 R τ T0 V2 P 1 T0 冮 T 0 1V22 dt 1 T0 V2 t冟0 τ v1t2 b V 0 t τ 0 τ 6 t T P 1 T0 冮 T02 T02 3v1t2 42 dt P lim TSx 1 T 冮 T2 T2 3f1t2 42 dt FIGURE 32 Binary digital signals: (a) τ/T 0.5; (b) τ/T 0.5 318
- 324. Digital Transmission QUESTIONS 1. Contrast the advantages and disadvantages of digital transmission. 2. What are the four most common methods of pulse modulation? 3. Which method listed in question 2 is the only form of pulse modulation that is used in a digital transmission system? Explain. 4. What is the purpose of the sample-and-hold circuit? 5. Define aperture and acquisition time. 6. What is the difference between natural and flat-top sampling? 7. Define droop. What causes it? 8. What is the Nyquist sampling rate? 9. Define and state the causes of foldover distortion. 10. Explain the difference between a magnitude-only code and a sign-magnitude code. 11. Explain overload distortion. 12. Explain quantizing. 13. What is quantization range? Quantization error? 14. Define dynamic range. 15. Explain the relationship between dynamic range, resolution, and the number of bits in a PCM code. 16. Explain coding efficiency. 17. What is SQR? What is the relationship between SQR, resolution, dynamic range, and the num- ber of bits in a PCM code? 18. Contrast linear and nonlinear PCM codes. 19. Explain idle channel noise. 20. Contrast midtread and midrise quantization. 21. Define companding. 22. What does the parameter μ determine? 23. Briefly explain the process of digital companding. 24. What is the effect of digital compression on SQR, resolution, quantization interval, and quanti- zation noise? 25. Contrast delta modulation PCM and standard PCM. 26. Define slope overload and granular noise. 27. What is the difference between adaptive delta modulation and conventional delta modulation? 28. Contrast differential and conventional PCM. PROBLEMS 1. Determine the Nyquist sample rate for a maximum analog input frequency of a. 4 kHz. b. 10 kHz. 2. For the sample-and-hold circuit shown in Figure 5a, determine the largest-value capacitor that can be used. Use the following parameters: an output impedance for Z1 20 Ω, an on resistance of Q1 of 20 Ω, an acquisition time of 10 μs, a maximum output current from Z1 of 20 mA, and an accuracy of 1%. 3. For a sample rate of 20 kHz, determine the maximum analog input frequency. 4. Determine the alias frequency for a 14-kHz sample rate and an analog input frequency of 8 kHz. 5. Determine the dynamic range for a 10-bit sign-magnitude PCM code. 319
- 325. 6. Determine the minimum number of bits required in a PCM code for a dynamic range of 80 dB. What is the coding efficiency? 7. For a resolution of 0.04 V, determine the voltages for the following linear seven-bit sign- magnitude PCM codes: a. 0 1 1 0 1 0 1 b. 0 0 0 0 0 1 1 c. 1 0 0 0 0 0 1 d. 0 1 1 1 1 1 1 e. 1 0 0 0 0 0 0 8. Determine the SQR for a 2-vrms signal and a quantization interval of 0.2 V. 9. Determine the resolution and quantization error for an eight-bit linear sign-magnitude PCM code for a maximum decoded voltage of 1.27 V. 10. A 12-bit linear PCM code is digitally compressed into eight bits. The resolution 0.03 V. De- termine the following for an analog input voltage of 1.465 V: a. 12-bit linear PCM code b. eight-bit compressed code c. Decoded 12-bit code d. Decoded voltage e. Percentage error 11. For a 12-bit linear PCM code with a resolution of 0.02 V, determine the voltage range that would be converted to the following PCM codes: a. 1 0 0 0 0 0 0 0 0 0 0 1 b. 0 0 0 0 0 0 0 0 0 0 0 0 c. 1 1 0 0 0 0 0 0 0 0 0 0 d. 0 1 0 0 0 0 0 0 0 0 0 0 e. 1 0 0 1 0 0 0 0 0 0 0 1 f. 1 0 1 0 1 0 1 0 1 0 1 0 12. For each of the following 12-bit linear PCM codes, determine the eight-bit compressed code to which they would be converted: a. 1 0 0 0 0 0 0 0 1 0 0 0 b. 1 0 0 0 0 0 0 0 1 0 0 1 c. 1 0 0 0 0 0 0 1 0 0 0 0 d. 0 0 0 0 0 0 1 0 0 0 0 0 e. 0 1 0 0 0 0 0 0 0 0 0 0 f. 0 1 0 0 0 0 1 0 0 0 0 0 13. Determine the Nyquist sampling rate for the following maximum analog input frequencies: 2 kHz, 5 kHz, 12 kHz, and 20 kHz. 14. For the sample-and-hold circuit shown in Figure 5a, determine the largest-value capacitor that can be used for the following parameters: Z1 output impedance 15 Ω, an on resistance of Q1 of 15 Ω, an acquisition time of 12 μs, a maximum output current from Z1 of 10 mA, an accuracy of 0.1%, and a maximum change in voltage dv 10 V. 15. Determine the maximum analog input frequency for the following Nyquist sample rates: 2.5 kHz, 4 kHz, 9 kHz, and 11 kHz. 16. Determine the alias frequency for the following sample rates and analog input frequencies: fa (kHz) fs (kHz) 3 4 5 8 6 8 5 7 17. Determine the dynamic range in dB for the following n-bit linear sign-magnitude PCM codes: n 7, 8, 12, and 14. 18. Determine the minimum number of bits required for PCM codes with the following dynamic ranges and determine the coding efficiencies: DR 24 dB, 48 dB, and 72 dB. Digital Transmission 320
- 326. 19. For the following values of μ, Vmax, and Vin, determine the compressor gain: μ Vmax (V) Vin (V) 255 1 0.75 100 1 0.75 255 2 0.5 20. For the following resolutions, determine the range of the eight-bit sign-magnitude PCM codes: Code Resolution (V) 10111000 0.1 00111000 0.1 11111111 0.05 00011100 0.02 00110101 0.02 11100000 0.02 00000111 0.02 21. Determine the SQR for the following input signal and quantization noise magnitudes: Vs Vn (V) 1 vrms 0.01 2 vrms 0.02 3 vrms 0.01 4 vrms 0.2 22. Determine the resolution and quantization noise for an eight-bit linear sign-magnitude PCM code for the following maximum decoded voltages: Vmax 3.06 Vp, 3.57 Vp, 4.08 Vp, and 4.59 Vp. 23. A 12-bit linear sign-magnitude PCM code is digitally compressed into 8 bits. For a resolution of 0.016 V, determine the following quantities for the indicated input voltages: 12-bit linear PCM code, eight-bit compressed code, decoded 12-bit code, decoded voltage, and percentage error. Vin 6.592 V, 12.992 V, and 3.36 V. 24. For the 12-bit linear PCM codes given, determine the voltage range that would be converted to them: 12-Bit Linear Code Resolution (V) 100011110010 0.12 000001000000 0.10 000111111000 0.14 111111110000 0.12 25. For the following 12-bit linear PCM codes, determine the eight-bit compressed code to which they would be converted: 12-Bit Linear Code 100011110010 000001000000 000111111000 111111110010 000000100000 26. For the following eight-bit compressed codes, determine the expanded 12-bit code. Eight-Bit Code 11001010 00010010 10101010 01010101 11110000 11011011 Digital Transmission 321
- 327. Digital Transmission ANSWERS TO SELECTED PROBLEMS 1. a. 8 kHz b. 20 kHz 3. 10 kHz 5. 6 kHz 7. a. 2.12 V b. 0.12 V c. 0.04 V d. 2.12 V e. 0 V 9. 1200 or 30.8 dB 11. a. 0.01 to 0.03 V b. 0.01 to 0.03 V c. 20.47 to 20.49 V d. 20.47 to 20.49 V e. 5.13 to 5.15 V f. 13.63 to 13.65 V 13. fin fs 2 kHz 4 kHz 5 kHz 10 kHz 12 kHz 24 kHz 20 kHz 40 kHz 15. fin fs 2.5 kHz 1.25 kHz 4 kHz 2 kHz 9 kHz 4.5 kHz 11 kHz 5.5 kHz 17. N DR db 7 63 6 8 127 12 12 2047 66 14 8191 78 19. μ gain 255 0.948 100 0.938 255 1.504 21. 50.8 dB, 50.8 dB, 60.34 dB, 36.82 dB 23. Vin 12-bit 8-bit 12-bit decoded V % Error 6.592 100110011100 11011001 100110011000 6.528 0.98 12.992 001100101100 01100010 001100101000 12.929 0.495 3.36 100011010010 11001010 100011010100 3.392 0.94 25. 11001110, 00110000, 01011111, 00100000 322
- 328. Digital T-Carriers and Multiplexing CHAPTER OUTLINE 1 Introduction 9 Bit versus Word Interleaving 2 Time-Division Multiplexing 10 Statistical Time-Division Multiplexing 3 T1 Digital Carrier 11 Codecs and Combo Chips 4 North American Digital Hierarchy 12 Frequency-Division Multiplexing 5 Digital Carrier Line Encoding 13 ATT’s FDM Hierarchy 6 T Carrier Systems 14 Composite Baseband Signal 7 European Digital Carrier System 15 Formation of a Mastergroup 8 Digital Carrier Frame Synchronization 16 Wavelength-Division Multiplexing OBJECTIVES ■ Define multiplexing ■ Describe the frame format and operation of the T1 digital carrier system ■ Describe the format of the North American Digital Hierarchy ■ Define line encoding ■ Define the following terms and describe how they affect line encoding: duty cycle, bandwidth, clock recovery, er- ror detection, and detecting and decoding ■ Describe the basic T carrier system formats ■ Describe the European digital carrier system ■ Describe several methods of achieving frame synchronization ■ Describe the difference between bit and word interleaving ■ Define codecs and combo chips and give a brief explanation of how they work ■ Define frequency-division multiplexing ■ Describe the format of the North American FDM Hierarchy ■ Define and describe baseband and composite baseband signals ■ Explain the formation of a mastergroup ■ Describe wavelength-division multiplexing ■ Explain the advantages and disadvantages of wavelength-division multiplexing From Chapter 7 of Advanced Electronic Communications Systems, Sixth Edition. Wayne Tomasi. Copyright © 2004 by Pearson Education, Inc. Published by Pearson Prentice Hall. All rights reserved. 323
- 329. 1 INTRODUCTION Multiplexing is the transmission of information (in any form) from one or more source to one or more destination over the same transmission medium (facility). Although transmis- sions occur on the same facility, they do not necessarily occur at the same time or occupy the same bandwidth. The transmission medium may be a metallic wire pair, a coaxial ca- ble, a PCS mobile telephone, a terrestrial microwave radio system, a satellite microwave system, or an optical fiber cable. There are several domains in which multiplexing can be accomplished, including space, phase, time, frequency, and wavelength. Space-division multiplexing (SDM) is a rather unsophisticated form of multiplexing that simply constitutes propagating signals from different sources on different cables that are contained within the same trench. The trench is considered to be the transmission medium. QPSK is a form of phase-division multiplexing (PDM) where two data channels (the I and Q) modulate the same carrier frequency that has been shifted 90° in phase. Thus, the I-channel bits modulate a sine wave carrier, while the Q-channel bits modulate a cosine wave carrier. After modulation has occurred, the I- and Q-channel carriers are linearly combined and propagated at the same time over the same transmission medium, which can be a cable or free space. The three most predominant methods of multiplexing signals are time-division mul- tiplexing (TDM), frequency-division multiplexing (FDM), and the more recently devel- oped wavelength-division multiplexing (WDM). The remainder of this chapter will be ded- icated to time-, frequency-, and wavelength-division multiplexing. 2 TIME-DIVISION MULTIPLEXING With time division multiplexing (TDM), transmissions from multiple sources occur on the same facility but not at the same time. Transmissions from various sources are interleaved in the time domain. PCM is the most prevalent encoding technique used for TDM digital signals. With a PCM-TDM system, two or more voice channels are sampled, converted to PCM codes, and then time-division multiplexed onto a single metallic or optical fiber cable. The fundamental building block for most TDM systems in the United States begins with a DS-0 channel (digital signal level 0). Figure 1 shows the simplified block diagram PAM Analog Parallel PCM code Input signal 0 to 4 kHz Antialiasing bandpass filter Sample pulse 8 kHz Conversion clock 1.536 MHz Line speed clock 64 kHz Analog-to digital converter Sample and hold Parallel-to serial converter 8-bit serial PCM code (64 kbps) DS-0 FIGURE 1 Single-channel (DS-O-level) PCM transmission system Digital T-Carriers and Multiplexing 324
- 330. for a DS-0 single-channel PCM system. As the figure shows, DS-0 channels use an 8-kHz sample rate and an eight-bit PCM code, which produces a 64-kbps PCM line speed: Figure 2a shows the simplified block diagram for a PCM carrier system comprised of two DS-0 channels that have been time-division multiplexed. Each channel’s input is sam- pled at an 8-kHz rate and then converted to an eight-bit PCM code. While the PCM code for channel 1 is being transmitted, channel 2 is sampled and converted to a PCM code. While the PCM code from channel 2 is being transmitted, the next sample is taken from channel 1 and converted to a PCM code. This process continues, and samples are taken al- ternately from each channel, converted to PCM codes, and transmitted. The multiplexer is simply an electronically controlled digital switch with two inputs and one output. Channel 1 and channel 2 are alternately selected and connected to the transmission line through the multiplexer. One eight-bit PCM code from each channel (16 total bits) is called a TDM frame, and the time it takes to transmit one TDM frame is called the frame time. The frame time is equal to the reciprocal of the sample rate (1/fs, or 1/8000 125 μs). Figure 2b shows the TDM frame allocation for a two-channel PCM system with an 8-kHz sample rate. The PCM code for each channel occupies a fixed time slot (epoch) within the total TDM frame. With a two-channel system, one sample is taken from each channel during each frame, and the time allocated to transmit the PCM bits from each channel is equal to one-half the total frame time. Therefore, eight bits from each channel must be transmitted during each frame (a total of 16 PCM bits per frame). Thus, the line speed at the output of the multiplexer is Although each channel is producing and transmitting only 64 kbps, the bits must be clocked out onto the line at a 128-kHz rate to allow eight bits from each channel to be transmitted in a 1211-μs time slot. 3 T1 DIGITAL CARRIER A digital carrier system is a communications system that uses digital pulse rather than ana- log signals to encode information. Figure 3a shows the block diagram for ATT’s T1 dig- ital carrier system, which has been the North American digital multiplexing standard since 1963 and recognized by the ITU-T as Recommendation G.733. T1 stands for transmission one and specifies a digital carrier system using PCM-encoded analog signals. A T1 carrier system time-division multiplexes PCM-encoded samples from 24 voice-band channels for transmission over a single metallic wire pair or optical fiber transmission line. Each voice- band channel has a bandwidth of approximately 300 Hz to 3000 Hz. Again, the multiplexer is simply a digital switch with 24 independent inputs and one time-division multiplexed output. The PCM output signals from the 24 voice-band channels are sequentially selected and connected through the multiplexer to the transmission line. Simply, time-division multiplexing 24 voice-band channels does not in itself consti- tute a T1 carrier system. At this point, the output of the multiplexer is simply a multiplexed first-level digital signal (DS level 1). The system does not become a T1 carrier until it is line encoded and placed on special conditioned cables called T1 lines. Line encoding is de- scribed later in this chapter. 2 channels frame 8000 frames second 8 bits channel 128 kbps 64,000 bps line speed 8000 samples second 8 bits sample Digital T-Carriers and Multiplexing 325
- 332. Digital T-Carriers and Multiplexing FIGURE 3 Bell system T1 digital carrier system: (a) block diagram; (b) sampling sequence 327
- 333. With a T1 carrier system, D-type (digital) channel banks perform the sampling, en- coding, and multiplexing of 24 voice-band channels. Each channel contains an eight-bit PCM code and is sampled 8000 times a second. Each channel is sampled at the same rate but not necessarily at the same time. Figure 3b shows the channel sampling sequence for a 24-channel T1 digital carrier system. As the figure shows, each channel is sampled once each frame but not at the same time. Each channel’s sample is offset from the previous chan- nel’s sample by 1/24 of the total frame time. Therefore, one 64-kbps PCM-encoded sample is transmitted for each voice-band channel during each frame (a frame time of 1/8000 125 μs). The line speed is calculated as follows: thus Later, an additional bit (called the framing bit) is added to each frame. The framing bit occurs once per frame (8000-bps rate) and is recovered in the receiver, where it is used to maintain frame and sample synchronization between the TDM transmitter and receiver. As a result, each frame contains 193 bits, and the line speed for a T1 digital carrier system is 3-1 D-Type Channel Banks Early T1 carrier systems used D1 digital channel banks (PCM encoders and decoders) with a seven-bit magnitude-only PCM code, analog companding, and μ 100. A later version of the D1 digital channel bank added an eighth bit (the signaling bit) to each PCM code for performing interoffice signaling (supervision between telephone offices, such as on hook, off hook, dial pulsing, and so forth). Since a signaling bit was added to each sample in every frame, the signaling rate was 8 kbps. In the early digital channel banks, the framing bit se- quence was simply an alternating 1/0 pattern. Figure 4 shows the frame and bit alignment for T1-carrier systems that used D1 channel banks. Over the years, T1 carrier systems have generically progressed through D2, D3, D4, D5, and D6 channel banks. D4, D5, and D6 channel banks use digital companding and eight-bit sign-magnitude-compressed PCM codes with μ 255. Because the early D1 channel banks used a magnitude-only PCM code, an error in the most significant bit of a PCM sample produced a decoded error equal to one-half the to- tal quantization range. Newer version digital channel banks used sign-magnitude codes, and an error in the sign bit causes a decoded error equal to twice the sample magnitude (V to V or vice versa) with a worst-case error equal to twice the total quantization range. However, in practice, maximum amplitude samples occur rarely; therefore, most errors have a magnitude less than half the coding range. On average, performance with sign-magnitude PCM codes is much better than with magnitude-only codes. 3-2 Superframe TDM Format The 8-kbps signaling rate used with the early digital channel banks was excessive for sig- naling on standard telephone voice circuits. Therefore, with modern channel banks, a sig- naling bit is substituted only into the least significant bit of every sixth frame. Hence, five of every six frames have eight-bit resolution, while one in every six frames (the signaling frame) has only seven-bit resolution. Consequently, the signaling rate on each channel is only 1.333 kbps (8000 bps/6), and the average number of bits per sample is actually 75/6 bits. Becauseonlyeverysixthframeincludesasignalingbit,itisnecessarythatalltheframes be numbered so that the receiver knows when to extract the signaling bit. Also, because the signaling is accomplished with a two-bit binary word, it is necessary to identify the most and least significant bits of the signaling word. Consequently, the superframe format shown in 193 bits frame 8000 frames second 1.544 Mbps 192 bits frame 8000 frames second 1.536 Mbps 24 channels frame 8 bits channel 192 bits per frame Digital T-Carriers and Multiplexing 328
- 334. (a) 1 TDM Frame Framing bit Sample 1 Sample 2 Channel 1 Channel 2 Channel 3 (7-bit magnitude-only PCM codes) b6 b5 b4 b3 b2 b1 b0 Channel 4-23 Channel 24 Channel 1 Framing bit Framing bit MSB LSB FIGURE 4 Early T1 Carrier system frame and sample alignment: (a) seven-bit magnitude-only PCM code; (b) seven-bit sign-magnitude code; (c) seven-bit sign-magnitude PCM code with signaling bit (b) 1 TDM Frame Framing bit Sample 1 Sample 2 Channel 1 Channel 2 Channel 3 (7-bit sign-magnitude PCM codes) b6 b5 b4 b3 b2 b1 b0 Channel 4-23 Channel 24 Channel 1 Framing bit Framing bit Sign bit MSB LSB (c) 1 TDM Frame Framing bit Sample 1 Sample 2 Channel 1 Channel 2 Channel 3 (7-bit sign-magnitude PCM codes with signaling bit) b7 b6 b5 b4 b3 b2 b1 Channel 4-23 Channel 24 Channel 1 Framing bit Framing bit Signal bit Sign bit MSB LSB b0 Digital T-Carriers and Multiplexing 329
- 335. FIGURE 5 Framing bit sequence for the T1 superframe format using D2 or D3 channel banks: (a) frame synchronizing bits (odd-numbered frames); (b) signaling frame alignment bits (even-numbered frames); (c) composite frame alignment Figure 5 was devised. Within each superframe, there are 12 consecutively numbered frames (1 to 12). The signaling bits are substituted in frames 6 and 12, the most significant bit into frame 6, and the least significant bit into frame 12. Frames 1 to 6 are called theA-highway, with frame6designatedtheA-channelsignalingframe.Frames7to12arecalledtheB-highway, with frame 12 designated the B-channel signaling frame. Therefore, in addition to identifying the signaling frames, the 6th and 12th frames must also be positively identified. To identify frames 6 and 12, a different framing bit sequence is used for the odd- and even-numbered frames. The odd frames (frames 1, 3, 5, 7, 9, and 11) have an alternating 1/0 pattern, and the even frames (frames 2, 4, 6, 8, 10, and 12) have a 0 0 1 1 1 0 repetitive pattern. As a result, the combined framing bit pattern is 1 0 0 0 1 1 0 1 1 1 0 0. The odd- numbered frames are used for frame and sample synchronization, and the even-numbered frames are used to identify theA- and B-channel signaling frames (frames 6 and 12). Frame 6 is identified by a 0/1 transition in the framing bit between frames 4 and 6. Frame 12 is identified by a 1/0 transition in the framing bit between frames 10 and 12. In addition to multiframe alignment bits and PCM sample bits, specific time slots are used to indicate alarm conditions. For example, in the case of a transmit power supply failure, a com- mon equipment failure, or loss of multiframe alignment, the second bit in each channel is made alogic0untilthealarmconditionhascleared.Also, theframingbitinframe12iscomplemented whenever multiframe alignment is lost, which is assumed whenever frame alignment is lost. In addition, there are special framing conditions that must be avoided to maintain clock and bit syn- chronization in the receive demultiplexing equipment. Figure 6 shows the frame, sample, and signaling alignment for the T1 carrier system using D2 or D3 channel banks. Digital T-Carriers and Multiplexing 330
- 337. Digital T-Carriers and Multiplexing Figure 7a shows the framing bit circuitry for the 24-channel T1 carrier system using D2 or D3 channel banks. Note that the bit rate at the output of the TDM multiplexer is 1.536 Mbps and that the bit rate at the output of the 193-bit shift register is 1.544 Mbps. The 8- kHz difference is due to the addition of the framing bit. D4 channel banks time-division multiplex 48 voice-band telephone channels and op- erate at a transmission rate of 3.152 Mbps. This is slightly more than twice the line speed for 24-channel D1, D2, or D3 channel banks because with D4 channel banks, rather than transmitting a single framing bit with each frame, a 10-bit frame synchronization pattern is used. Consequently, the total number of bits in a D4 (DS-1C) TDM frame is and the line speed for DS-1C systems is The framing for DS-1 (T1) PCM-TDM system or the framing pattern for the DS-1C (T1C) time-division multiplexed carrier system is added to the multiplexed digital signal at the output of the multiplexer. The framing bit circuitry used for the 48-channel DS-1C is shown in Figure 7b. 3-3 Extended Superframe Format Another framing format recently developed for new designs of T1 carrier systems is the ex- tended superframe format. The extended superframe format consists of 24 193-bit frames, totaling 4632 bits, of which 24 are framing bits. One extended superframe occupies 3 ms: A framing bit occurs once every 193 bits; however, only 6 of the 24 framing bits are used for frame synchronization. Frame synchronization bits occur in frames 4, 8, 12, 16, 20, and 24 and have a bit sequence of 0 0 1 0 1 1. Six additional framing bits in frames 1, 5, 9, 13, 17, and 21 are used for an error detection code called CRC-6 (cyclic redundancy checking). The 12 remaining framing bits provide for a management channel called the facilities data link (FDL). FDL bits occur in frames 2, 3, 6, 7, 10, 11, 14, 15, 18, 19, 22, and 23. The extended superframe format supports a four-bit signaling word with signaling bits provided in the second least significant bit of each channel during every sixth frame. The sig- naling bit in frame 6 is called theA bit, the signaling bit in frame 12 is called the B bit, the sig- naling bit in frame 18 is called the C bit, and the signaling bit in frame 24 is called the D bit. These signaling bit streams are sometimes called theA, B, C, and D signaling channels (or sig- naling highways). The extended superframe framing bit pattern is summarized in Table 1. 3-4 Fractional T Carrier Service Fractional T carrier emerged because standard T1 carriers provide a higher capacity (i.e., higher bit rate) than most users require. Fractional T1 systems distribute the channels (i.e., bits) in a standard T1 system among more than one user, allowing several subscribers to share one T1 line. For example, several small businesses located in the same building can share one T1 line (both its capacity and its cost). Bit rates offered with fractional T1 carrier systems are 64 kbps (1 channel), 128 kbps (2 channels), 256 kbps (4 channels), 384 kbps (6 channels), 512 kbps (8 channels), and 768 kbps (12 channels) with 384 kbps (1/4 T1) and 768 kbps (1/2 T1) being the most common. The minimum data rate necessary to propagate video information is 384 kbps. ¢ 1 1.544 Mbitss ≤¢ 193 bits frame ≤ 124 frames2 3 ms 394 bits frame 8000 frames second 3.152 Mbps 8 bits channel 48 channels frame 384 bits frame 10 framing bits frame 394 bits frame 332
- 338. Digital T-Carriers and Multiplexing Buffer channel 1 Buffer channel 24 Channel 1 input PCM encoder T i m e d i v i s i o n m u l t i p l e x e r 1.536 MHz clock 8 kHz clock clock 1.544 MHz 1.544 Mbps PCM TDM FB Framing bit circuit 1.536 MHz clock 192-bit shift register Bits 2-191 193-bit shift register 192 1 1 192 1.536 Mbps 1.536 MHz clock 64 kbps 64 kbps Channel 24 input PCM encoder Channels 2–23 (a) FIGURE 7 Framing bit circuitry T1 carrier system: (a) DS-1; (b) DS-1C Channel 1 input PCM encoder 3.072 MHz clock 8-bit PCM code 64 kbps 3.072 Mbps 384-bit shift register 3.152 Mbps PCM-TDM Bits 2–383 384 1 Framing word 10-bits 384 PCM bits sync 8-bit PCM code 64 kbps Channel 48 input PCM encoder Buffer channel 1 Buffer channel 48 Channels 2–47 inputs 3.072 MHz clock (b) Time- division multiplexer Buffers channels 2-47 1 394-bit shift register 385 394 333
- 339. Digital T-Carriers and Multiplexing Table 1 Extended Superframe Format Frame Number Framing Bit Frame Number Framing Bit 1 C 13 C 2 F 14 F 3 F 15 F 4 S 0 16 S 0 5 C 17 C 6 F 18 F 7 F 19 F 8 S 0 20 S 1 9 C 21 C 10 F 22 F 11 F 23 F 12 S 1 24 S 1 1.544 Mbps T1 line (Users 1, 2, 3, and 4) User 1 128 kbps 256 kbps 384 kbps 8 kbps framing bits 768 kbps User 2 User 3 User 4 DSU/CSU FIGURE 8 Fractional T1 carrier service Fractional T3 is essentially the same as fractional T1 except with higher channel capacities, higher bit rates, and more customer options. Figure 8 shows four subscribers combining their transmissions in a special unit called a data service unit/channel service unit (DSU/CSU). A DSU/CSU is a digital interface that provides the physical connection to a digital carrier network. User 1 is allocated 128 kbps, user 2 256 kbps, user 3 384 kbps, and user 4 768 kbps, for a total of 1.536 kbps (8 kbps is reserved for the framing bit). 4 NORTH AMERICAN DIGITAL HIERARCHY Multiplexing signals in digital form lends itself easily to interconnecting digital transmis- sion facilities with different transmission bit rates. Figure 9 shows the American Telephone and Telegraph Company (ATT) North American Digital Hierarchy for multiplexing dig- ital signals from multiple sources into a single higher-speed pulse stream suitable for transmission on the next higher level of the hierarchy. To upgrade from one level in the 334
- 341. Digital T-Carriers and Multiplexing Table 2 North American Digital Hierarchy Summary Channel Line Type Digital Signal Bit Rate Capacities Services Offered T1 DS-1 1.544 Mbps 24 Voice-band telephone or data Fractional T1 DS-1 64 kbps to 1.536 Mbps 24 Voice-band telephone or data T1C DS-1C 3.152 Mbps 48 Voice-band telephone or data T2 DS-2 6.312 Mbps 96 Voice-band telephone, data, or picture phone T3 DS-3 44.736 Mbps 672 Voice-band telephone, data, picture phone, and broadcast-quality television Fractional T3 DS-3 64 kbps to 23.152 Mbps 672 Voice-band telephone, data, picture phone, and broadcast-quality television T4M DS-4 274.176 Mbps 4032 Same as T3 except more capacity T5 DS-5 560.160 Mbps 8064 Same as T3 except more capacity hierarchy to the next higher level, a special device called muldem (multiplexers/demultiplexer) is required. Muldems can handle bit-rate conversions in both directions. The muldem des- ignations (M112, M23, and so on) identify the input and output digital signals associated with that muldem. For instance, an M12 muldem interfaces DS-1 and DS-2 digital signals. An M23 muldem interfaces DS-2 and DS-3 digital signals. As the figure shows, DS-1 sig- nals may be further multiplexed or line encoded and placed on specially conditioned cables called T1 lines. DS-2, DS-3, DS-4, and DS-5 digital signals may be placed on T2, T3, T4M, or T5 lines, respectively. Digital signals are routed at central locations called digital cross-connects. A digital cross-connect (DSX) provides a convenient place to make patchable interconnects and per- form routine maintenance and troubleshooting. Each type of digital signal (DS-1, DS-2, and so on) has its own digital switch (DSX-1, DSX-2, and so on). The output from a digital switch may be upgraded to the next higher level of multiplexing or line encoded and placed on its respective T lines (T1, T2, and so on). Table 2 lists the digital signals, their bit rates, channel capacities, and services offered for the line types included in the North American Digital Hierarchy. 4-1 Mastergroup and Commercial Television Terminals Figure 10 shows the block diagram of a mastergroup and commercial television terminal. The mastergroup terminal receives voice-band channels that have already been frequency-di- vision multiplexed (a topic covered later in this chapter) without requiring that each voice- band channel be demultiplexed to voice frequencies. The signal processor provides fre- quency shifting for the mastergroup signals (shifts it from a 564-kHz to 3084-kHz bandwidth to a 0-kHz to 2520-kHz bandwidth) and dc restoration for the television signal. By shifting the mastergroup band, it is possible to sample at a 5.1-MHz rate. Sampling of the commercial television signal is at twice that rate or 10.2 MHz. When the bandwidth of the signals to be transmitted is such that after digital conver- sion it occupies the entire capacity of a digital transmission line, a single-channel terminal is provided. Examples of such single-channel terminals are mastergroup, commercial tele- vision, and picturephone terminals. 336
- 342. Digital T-Carriers and Multiplexing FIGURE 10 Block diagram of a mastergroup or commercial television digital terminal To meet the transmission requirements, a nine-bit PCM code is used to digitize each sample of the mastergroup or television signal. The digital output from the terminal is, therefore, approximately 46 Mbps for the mastergroup and twice that much (92 Mbps) for the television signal. The digital terminal shown in Figure 10 has three specific functions: (1) It converts the parallel data from the output of the encoder to serial data, (2) it inserts frame synchro- nizing bits, and (3) it converts the serial binary signal to a form more suitable for transmis- sion. In addition, for the commercial television terminal, the 92-Mbps digital signal must be split into two 46-Mbps digital signals because there is no 92-Mbps line speed in the digital hierarchy. 4-2 Picturephone Terminal Essentially, picturephone is a low-quality video transmission for use between nondedi- cated subscribers. For economic reasons, it is desirable to encode a picturephone signal into the T2 capacity of 6.312 Mbps, which is substantially less than that for commercial network broadcast signals. This substantially reduces the cost and makes the service af- fordable. At the same time, it permits the transmission of adequate detail and contrast resolution to satisfy the average picturephone subscriber. Picturephone service is ideally suited to a differential PCM code. Differential PCM is similar to conventional PCM ex- cept that the exact magnitude of a sample is not transmitted. Instead, only the difference between that sample and the previous sample is encoded and transmitted. To encode the difference between samples requires substantially fewer bits than encoding the actual sample. 4-3 Data Terminal The portion of communications traffic that involves data (signals other than voice) is in- creasing exponentially. Also, in most cases the data rates generated by each individual sub- scriber are substantially less than the data rate capacities of digital lines. Therefore, it seems only logical that terminals be designed that transmit data signals from several sources over the same digital line. Data signals could be sampled directly; however, this would require excessively high sample rates, resulting in excessively high transmission bit rates, especially for sequences 337
- 343. FIGURE 11 Data coding format of data with few or no transitions. A more efficient method is one that codes the transition times. Such a method is shown in Figure 11. With the coding format shown, a three-bit code is used to identify when transitions occur in the data and whether that transition is from a 1 to a 0 or vice versa. The first bit of the code is called the address bit. When this bit is a logic 1, this indicates that no transition occurred; a logic 0 indicates that a transi- tion did occur. The second bit indicates whether the transition occurred during the first half (0) or during the second half (1) of the sample interval. The third bit indicates the sign or direction of the transition; a 1 for this bit indicates a 0-to-1 transition, and a 0 in- dicates a 1-to-0 transition. Consequently, when there are no transitions in the data, a sig- nal of all 1s is transmitted. Transmission of only the address bit would be sufficient; how- ever, the sign bit provides a degree of error protection and limits error propagation (when one error leads to a second error and so on). The efficiency of this format is approximately 33%; there are three code bits for each data bit. The advantage of using a coded format rather than the original data is that coded data are more efficiently substituted for voice in analog systems. Without this coding format, transmitting a 250-kbps data signal requires the same bandwidth as would be required to transmit 60 voice channels with analog multiplexing. With this coded format, a 50-kbps data signal displaces three 64-kbps PCM-encoded channels, and a 250-kbps data stream displaces only 12 voice- band channels. 5 DIGITAL CARRIER LINE ENCODING Digital line encoding involves converting standard logic levels (TTL, CMOS, and the like) to a form more suitable to telephone line transmission. Essentially, six primary factors must be considered when selecting a line-encoding format: 1. Transmission voltages and DC component 2. Duty cycle 3. Bandwidth considerations 4. Clock and framing bit recovery 5. Error detection 6. Ease of detection and decoding Digital T-Carriers and Multiplexing 338
- 344. 5-1 Transmission Voltages and DC Component Transmission voltages or levels can be categorized as being either unipolar (UP) or bipolar (BP). Unipolar transmission of binary data involves the transmission of only a single nonzero voltage level (e.g., either a positive or a negative voltage for a logic 1 and 0 V [ground] for a logic 0). In bipolar transmission, two nonzero voltages are involved (e.g., a positive voltage for a logic 1 and an equal-magnitude negative voltage for a logic 0 or vice versa). Over a digital transmission line, it is more power efficient to encode binary data with voltages that are equal in magnitude but opposite in polarity and symmetrically balanced about 0V. For example, assuming a 1-ohm resistance and a logic 1 level of 5V and a logic 0 level of 0 V, the average power required is 12.5 W, assuming an equal probability of the occurrence of a logic 1 or a logic 0. With a logic 1 level of 2.5 V and a logic 0 level of 2.5 V, the average power is only 6.25 W. Thus, by using bipolar symmetrical voltages, the average power is reduced by a factor of 50%. 5-2 Duty Cycle The duty cycle of a binary pulse can be used to categorize the type of transmission. If the bi- nary pulse is maintained for the entire bit time, this is called nonreturn to zero (NRZ). If the ac- tive time of the binary pulse is less than 100% of the bit time, this is called return to zero (RZ). Unipolar and bipolar transmission voltages can be combined with either RZ or NRZ in several ways to achieve a particular line-encoding scheme. Figure 12 shows five line-en- coding possibilities. In Figure 12a, there is only one nonzero voltage level (V logic 1); a zero voltage indicates a logic 0. Also, each logic 1 condition maintains the positive voltage for the en- tire bit time (100% duty cycle). Consequently, Figure 12a represents a unipolar nonre- turn-to-zero signal (UPNRZ). Assuming an equal number of 1s and 0s, the average dc volt- age of a UPNRZ waveform is equal to half the nonzero voltage (V/2). In Figure 12b, there are two nonzero voltages (V logic 1 and V logic 0) and a 100% duty cycle is used. Therefore, Figure 12b represents a bipolar nonreturn-to-zero sig- nal (BPNRZ). When equal-magnitude voltages are used for logic 1s and logic 0s, and as- suming an equal probability of logic 1s and logic 0s occurring, the average dc voltage of a BPNRZ waveform is 0 V. In Figure 12c, only one nonzero voltage is used, but each pulse is active for only 50% of a bit time (tb/2). Consequently, the waveform shown in Figure 12c represents a unipolar return-to-zero signal (UPRZ).Assuming an equal probability of 1s and 0s occurring, the av- erage dc voltage of a UPRZ waveform is one-fourth the nonzero voltage (V/4). Figure 12d shows a waveform where there are two nonzero voltages (V logic 1 and V logic 0). Also, each pulse is active only 50% of a bit time. Consequently, the waveform shown in Figure 8d represents a bipolar return-to-zero (BPRZ) signal. Assuming equal-magnitude voltages for logic 1s and logic 0s and an equal probability of 1s and 0s oc- curring, the average dc voltage of a BPRZ waveform is 0 V. In Figure 12e, there are again two nonzero voltage levels (V and V), but now both polarities represent logic 1s, and 0 V represents a logic 0. This method of line encoding is called alternate mark inversion (AMI). With AMI transmissions, successive logic 1s are in- verted in polarity from the previous logic 1. Because return to zero is used, the encoding tech- nique is called bipolar-return-to-zero alternate mark inversion (BPRZ-AMI). The average dc voltage of a BPRZ-AMI waveform is approximately 0V regardless of the bit sequence. With NRZ encoding, a long string of either logic 1s or logic 0s produces a condition in which a receive may lose its amplitude reference for optimum discrimination between received 1s and 0s. This condition is called dc wandering. The problem may also arise when there is a significant imbalance in the number of 1s and 0s transmitted. Figure 13 shows how dc wandering is produced from a long string of successive logic 1s. It can be seen that after a long string of 1s, 1-to-0 errors are more likely than 0-to-1 errors. Similarly, long strings of logic 0s increase the probability of 0-to-1 errors. Digital T-Carriers and Multiplexing 339
- 345. FIGURE 12 Line-encoding formats: (a) UPNRZ; (b) BPNRZ; (c) UPRZ; (d) BPRZ; (e) BPRZ-AMI FIGURE 13 DC wandering The method of line encoding used determines the minimum bandwidth required for transmission, how easily a clock may be extracted from it, how easily it may be decoded, the average dc voltage level, and whether it offers a convenient means of detecting errors. 5-3 Bandwidth Requirements To determine the minimum bandwidth required to propagate a line-encoded digital signal, you must determine the highest fundamental frequency associated with the signal (see Figure 12). The highest fundamental frequency is determined from the worst-case (fastest transi- tion) binary bit sequence. With UPNRZ, the worst-case condition is an alternating 1/0 se- quence; the period of the highest fundamental frequency takes the time of two bits and, there- fore, is equal to one-half the bit rate (fb/2).With BPNRZ, again the worst-case condition is an Digital T-Carriers and Multiplexing 340
- 346. Table 3 Line-Encoding Summary Encoding Minimum Average Clock Error Format BW DC Recovery Detection UPNRZ fb/2* V/2 Poor No BPNRZ fb/2* 0 V* Poor No UPRZ fb V/4 Good No BPRZ fb 0 V* Best* No BPRZ-AMI fb/2* 0 V* Good Yes* * Denotes best performance or quality. alternating 1/0 sequence, and the highest fundamental frequency is one-half the bit rate (fb/2). With UPRZ, the worst-case condition occurs when two successive logic 1s occur. Therefore, the minimum bandwidth is equal to the bit rate (fb).With BPRZ encoding, the worst-case con- dition occurs for successive logic 1s or successive logic 0s, and the minimum bandwidth is again equal to the bit rate (fb).With BPRZ-AMI, the worst-case condition is two or more con- secutive logic 1s, and the minimum bandwidth is equal to one-half the bit rate (fb /2). 5-4 Clock and Framing Bit Recovery To recover and maintain clock and framing bit synchronization from the received data, there must be sufficient transitions in the data waveform. With UPNRZ and BPNRZ encoding, a long string of 1s or 0s generates a data signal void of transitions and, therefore, is inade- quate for clock recovery. With UPRZ and BPRZ-AMI encoding, a long string of 0s also generates a data signal void of transitions. With BPRZ, a transition occurs in each bit posi- tion regardless of whether the bit is a 1 or a 0. Thus, BPRZ is the best encoding scheme for clock recovery. If long sequences of 0s are prevented from occurring, BPRZ-AMI encod- ing provides sufficient transitions to ensure clock synchronization. 5-5 Error Detection With UPNRZ, BPNRZ, UPRZ, and BPRZ encoding, there is no way to determine if the re- ceived data have errors. However, with BPRZ-AMI encoding, an error in any bit will cause a bipolar violation (BPV, or the reception of two or more consecutive logic 1s with the same polarity). Therefore, BPRZ-AMI has a built-in error-detection mechanism. T carriers use BPRZ-AMI with 3 V and 3 V representing a logic 1 and 0 V representing a logic 0. Table 3 summarizes the bandwidth, average voltage, clock recovery, and error-de- tection capabilities of the line-encoding formats shown in Figure 12. From Table 3, it can be seen that BPRZ-AMI encoding has the best overall characteristics and is, therefore, the most commonly used encoding format. 5-6 Digital Biphase Digital biphase (sometimes called the Manchester code or diphase) is a popular type of line encoding that produces a strong timing component for clock recovery and does not cause dc wandering. Biphase is a form of BPRZ encoding that uses one cycle of a square wave at 0° phase to represent a logic 1 and one cycle of a square wave at 180° phase to represent a logic 0. Digital biphase encoding is shown in Figure 14. Note that a transition occurs in the center of every signaling element regardless of its logic condition or phase. Thus, biphase produces a strong timing component for clock recovery. In addition, assuming an equal probability of 1s and 0s, the average dc voltage is 0 V, and there is no dc wandering. A dis- advantage of biphase is that it contains no means of error detection. Biphaseencodingschemeshaveseveralvariations,includingbiphaseM,biphaseL,and biphase S. Biphase M is used for encoding SMPTE (Society of Motion Picture andTelevision Engineers) time-code data for recording on videotapes. Biphase M is well suited for this ap- plication because it has no dc component, and the code is self-synchronizing (self-clock- ing). Self-synchronization is an import feature because it allows clock recovery from the Digital T-Carriers and Multiplexing 341
- 347. FIGURE 14 Digital biphase FIGURE 15 Biphase, Miller, and dicode encoding formats data stream even when the speed varies with tape speed, such as when searching through a tape ineitherthefastortheslowmode.BiphaseLiscommonlycalledtheManchestercode.Biphase L is specified in IEEE standard 802.3 for Ethernet local area networks. Miller codes are forms of delay-modulated codes where a logic 1 condition produces a transition in the middle of the clock pulse, and a logic 0 produces no transition at the end of the clock intervals unless followed by another logic 0. Dicodes are multilevel binary codes that use more than two voltage levels to repre- sent the data. Bipolar RZ and RZ-AMI are two dicode encoding formats already discussed. Dicode NRZ and dicode RZ are two more commonly used dicode formats. Figure 15 shows several variations of biphase, Miller, and dicode encoding, and Table 4 summarizes their characteristics. 6 T CARRIER SYSTEMS T carriers are used for the transmission of PCM-encoded time-division multiplexed digital signals. In addition, T carriers utilize special line-encoded signals and metallic cables that have been conditioned to meet the relatively high bandwidths required for high-speed dig- ital transmission. Digital signals deteriorate as they propagate along a cable because of power losses in the metallic conductors and the low-pass filtering inherent in parallel-wire transmission lines. Consequently, regenerative repeaters must be placed at periodic inter- vals. The distance between repeaters depends on the transmission bit rate and the line- encoding technique used. Digital T-Carriers and Multiplexing 342
- 348. Table 4 Summary of Biphase, Miller, and Dicode Encoding Formats Biphase M (biphase-mark) 1 (hi)—transition in the middle of the clock interval 0 (low)—no transition in the middle of the clock interval Note: There is always a transition at the beginning of the clock interval. Biphase L (biphase-level/Manchester) 1 (hi)—transition from high to low in the middle of the clock interval 0 (low)—transition from low to high in the middle of the clock interval Biphase S (biphase-space) 1 (hi)—no transition in the middle of the clock interval 0 (low)—transition in the middle of the clock interval Note: There is always a transition at the beginning of the clock interval. Differential Manchester 1 (hi)—transition in the middle of the clock interval 0 (low)—transition at the beginning of the clock interval Miller/delay modulation 1 (hi)—transition in the middle of the clock interval 0 (low)—no transition at the end of the clock interval unless followed by a zero Dicode NRZ One-to-zero and zero-to-one data transitions change the signal polarity. If the data remain constant, then a zero-voltage level is output. Dicode RZ One-to-zero and zero-to-one data transitions change the signal polarity in half-step voltage increments. If the data do not change, then a zero-voltage level is output. FIGURE 16 Regenerative repeater block diagram Figure 16 shows the block diagram for a regenerative repeater. Essentially, there are three functional blocks: an amplifier/equalizer, a timing clock recovery circuit, and the reg- enerator itself. The amplifier/equalizer filters and shapes the incoming digital signal and raises its power level so that the regenerator circuit can make a pulse–no pulse decision. The timing clock recovery circuit reproduces the clocking information from the received data and provides the proper timing information to the regenerator so that samples can be made at the optimum time, minimizing the chance of an error occurring. A regenerative repeater is simply a threshold detector that compares the sampled voltage received to a reference level and determines whether the bit is a logic 1 or a logic 0. Spacing of repeaters is designed to maintain an adequate signal-to-noise ratio for error-free performance. The signal-to-noise ratio at the output of a regenerative repeater is Digital T-Carriers and Multiplexing 343
- 349. MSB LSB MSB LSB 1000 0000 0000 0001 Original DS-1 signal Original DS-1 signal 14 consecutive 0s (no substitution) MSB LSB MSB LSB 1000 0000 0000 0000 1000 0000 0000 0010 15 consecutive 0s Substituted DS-1 signal Substituted bit exactly what it was at the output of the transmit terminal or at the output of the previous re- generator (i.e., the signal-to-noise ratio does not deteriorate as a digital signal propagates through a regenerator; in fact, a regenerator reconstructs the original pulses with the origi- nal signal-to-noise ratio). 6-1 T1 Carrier Systems T1 carrier systems were designed to combine PCM and TDM techniques for short-haul transmission of 24 64-kbps channels with each channel capable of carrying digitally en- coded voice-band telephone signals or data. The transmission bit rate (line speed) for a T1 carrier is 1.544 Mbps, including an 8-kbps framing bit. The lengths of T1 carrier systems typically range from about 1 mile to over 50 miles. T1 carriers use BPRZ-AMI encoding with regenerative repeaters placed every 3000, 6000, or 9000 feet. These distances were selected because they were the distances between telephone company manholes where regenerative repeaters are placed. The transmission medium for T1 carriers is generally 19- to 22-gauge twisted-pair metallic cable. Because T1 carriers use BPRZ-AMI encoding, they are susceptible to losing clock synchronization on long strings of consecutive logic 0s. With a folded binary PCM code, the possibility of generating a long string of contiguous logic 0s is high. When a channel is idle, it generates a 0-V code, which is either seven or eight consecutive logic zeros. Therefore, whenever two or more adjacent channels are idle, there is a high likelihood that a long string of consecutive logic 0s will be transmitted. To reduce the possibility of transmitting a long string of consecutive logic 0s, the PCM data were complemented prior to transmission and then complemented again in the receiver before decoding. Consequently, the only time a long string of consecutive logic 0s are transmitted is when two or more adjacent channels each encode the maximum possible positive sample voltage, which is unlikely to happen. Ensuring that sufficient transitions occur in the data stream is sometimes called ones density. Early T1 and T1C carrier systems provided measures to ensure that no single eight- bit byte was transmitted without at least one bit being a logic 1 or that 15 or more consec- utive logic 0s were not transmitted. The transmissions from each frame are monitored for the presence of either 15 consecutive logic 0s or any one PCM sample (eight bits) without at least one nonzero bit. If either of these conditions occurred, a logic 1 is substituted into the appropriate bit position. The worst-case conditions were Digital T-Carriers and Multiplexing A 1 is substituted into the second least significant bit, which introduces an encoding error equal to twice the amplitude resolution. This bit is selected rather than the least sig- nificant bit because, with the superframe format, during every sixth frame the LSB is the signaling bit, and to alter it would alter the signaling word. If at any time 32 consecutive logic 0s are received, it is assumed that the system is not generating pulses and is, therefore, out of service. WithmodernT1carriers,atechniquecalledbinaryeightzerosubstitution(B8ZS)isused to ensure that sufficient transitions occur in the data to maintain clock synchronization. With 344
- 350. MSB LSB MSB MSB LSB LSB 1010 1000 0000 0000 0000 0001 1010 1000 0000 0010 0000 0001 Original DS-1 signal Substituted DS-1 signal Substituted bit B8ZS, whenever eight consecutive 0s are encountered, one of two special patterns is substi- tuted for the eight 0s, either 0 0 0 0 or 0 0 0 0. The (plus) and (mi- nus)representbipolarlogic1conditions,anda0(zero)indicatesalogic0condition.Theeight- bit pattern substituted for the eight consecutive 0s is the one that purposely induces bipolar violations in the fourth and seventh bit positions. Ideally, the receiver will detect the bipolar vi- olations and the substituted pattern and then substitute the eight 0s back into the data signal. During periods of low usage, eight logic 1s are substituted into idle channels.Two examples of Digital T-Carriers and Multiplexing Bit positions 8 0 + 7 0 8 zeros Substituted pattern 6 0 0 5 0 4 0 3 0 2 0 1 0 + 0 + 0 0 0 Original data Original waveform + 0 Substituted waveform Bipolar violations Bit positions 8 0 + 7 0 8 zeros Substituted pattern 6 0 0 5 0 4 0 3 0 2 0 1 0 0 + + 0 0 0 Original data Original waveform + 0 Substituted waveform Bipolar violations (a) (b) FIGURE 17 Waveforms for B8ZS example: (a) substitution pattern 1; (b) substitution pattern 2 345
- 351. Digital T-Carriers and Multiplexing 6-3 T3 Carrier System T3 carriers time-division multiplex 672 64-kbps voice or data channels for transmission over a single 3A-RDS coaxial cable. The transmission bit rate for T3 signals is 44.736 Mbps. The coding technique used with T3 carriers is binary three zero substitution (B3ZS). 6-2 T2 Carrier System T2 carriers time-division multiplex 96 64-kbps voice or data channels into a single 6.312- Mbps data signal for transmission over twisted-pair copper wire up to 500 miles over a spe- cial LOCAP (low capacitance) metallic cable. T2 carriers also use BPRZ-AMI encoding; however, because of the higher transmission rate, clock synchronization is even more crit- ical than with a T1 carrier. A sequence of six consecutive logic 0s could be sufficient to cause loss of clock synchronization. Therefore, T2 carrier systems use an alternative method of ensuring that ample transitions occur in the data. This method is called binary six zero substitution (B6ZS). With B6ZS, whenever six consecutive logic 0s occur, one of the following binary codes is substituted in its place: 0 0 or 0 0 . Again and represent logic 1s, and 0 represents a logic 0. The six-bit code substituted for the six consecutive 0s is selected to purposely cause a bipolar violation. If the violation is detected in the receiver, the original six 0s can be substituted back into the data signal. The substituted patterns produce bipolar vio- lations(i.e.,consecutivepulseswiththesamepolarity)inthesecondandfourthbitsofthesub- stituted patterns. If DS-2 signals are multiplexed to form DS-3 signals, the B6ZS code must be detected and removed from the DS-2 signal prior to DS-3 multiplexing. An example of B6ZS is illustrated here and its corresponding waveform shown in Figure 18. Bit positions Bipolar violation same polarity pulses MSB LSB 8 0 + 7 0 6 0 0 5 0 4 0 + 3 0 0 2 0 0 1 0 0 0 0 + + 0 0 + + 0 0 0 0 0 0 Original data Data with substitution Bit positions Original data Data with substitution Bipolar violation same polarity pulses MSB LSB 8 0 7 0 + 6 0 0 5 0 + 4 0 3 0 0 2 0 0 1 0 0 0 0 + + 0 0 0 0 0 0 0 0 Bit positions Bipolar violation same polarity pulses + + 0 0 6 0 0 5 0 0 4 3 0 0 2 + + 1 0 0 0 0 0 0 + 0 0 0 + 0 + + Original data Data with substitution B8ZS are illustrated here and their corresponding waveforms shown in Figures 17a and b, respectively: 346
- 352. Digital T-Carriers and Multiplexing Bit positions + + 0 6 zeros Substituted pattern 6 0 0 5 0 4 3 0 2 + 1 0 0 0 0 0 0 0 + 0 + 0 + Original data Original waveform + 0 Substituted waveform Bipolar violations—polarity of 2nd and 4th pulses of the substituted patterns are same FIGURE 18 Waveform for B6ZS example MSB LSB MSB MSB LSB LSB 0+00 000 +000 000 +00 00+0 Original data Bipolar violations Substituted patterns 3 0s 3 0s 3 0s 0+00 +00 +0 000 +00 0+0 Encoded data Substitutions are made for any occurrence of three consecutive 0s. There are four substitu- tion patterns used: 00, 0, 00, and 0. The pattern chosen should cause a bipolar error in the third substituted bit. An example of B3ZS is shown here: 6-4 T4M and T5 Carrier Systems T4M carriers time-division multiplex 4032 64-kbps voice or data channels for transmission over a single T4M coaxial cable up to 500 miles. The transmission rate is sufficiently high that substitute patterns are impractical. Instead, T4M carriers transmit scrambled unipolar NRZ digital signals; the scrambling and descrambling functions are performed in the sub- scriber’s terminal equipment. T5 carriers time-division multiplex 8064 64-kbps voice or data channels and transmit them at a 560.16 Mbps rate over a single coaxial cable. 7 EUROPEAN DIGITAL CARRIER SYSTEM In Europe, a different version of T carrier lines is used, called E-lines.Although the two sys- tems are conceptually the same, they have different capabilities. Figure 19 shows the frame alignment for the E1 European standard PCM-TDM system. With the basic E1 system, a 125-μs frame is divided into 32 equal time slots. Time slot 0 is used for a frame alignment pattern and for an alarm channel. Time slot 17 is used for a common signaling channel (CSC). The signaling for all 30 voice-band channels is accomplished on the common signaling channel. Consequently, 30 voice-band channels are time-division multiplexed into each E1 frame. 347
- 353. FIGURE 19 CCITT TDM frame alignment and common signaling channel alignment: (a) CCITT TDM frame (125 μs, 256 bits, 2.048 Mbps); (b) common signaling channel Table 5 European Transmission Rates and Capacities Transmission Channel Line Bit Rate (Mbps) Capacity E1 2.048 30 E2 8.448 120 E3 34.368 480 E4 139.264 1920 With the European E1 standard, each time slot has eight bits. Consequently, the total number of bits per frame is and the line speed for an E-1 TDM system is The European digital transmission system has aTDM multiplexing hierarchy similar to the North American hierarchy except the European system is based on the 32-time-slot (30- voice-channel) E1 system. The European Digital Multiplexing Hierarchy is shown in Table 5. Interconnecting T carriers with E carriers is not generally a problem because most multiplexers and demultiplexers are designed to perform the necessary bit rate conversions. 256 bits frame 8000 frames second 2.048 Mbps 8 bits time slot 32 time slots frame 256 bits frame Digital T-Carriers and Multiplexing 348
- 354. Digital T-Carriers and Multiplexing 8 DIGITAL CARRIER FRAME SYNCHRONIZATION With TDM systems, it is imperative not only that a frame be identified but also that indi- vidual time slots (samples) within the frame be identified. To acquire frame synchroniza- tion, a certain amount of overhead must be added to the transmission. There are several methods used to establish and maintain frame synchronization, including added-digit, robbed-digit, added-channel, statistical, and unique-line code framing. 8-1 Added-Digit Framing T1 carriers using D1, D2, or D3 channel banks use added-digit framing. A special framing digit (framing pulse) is added to each frame. Consequently, for an 8-kHz sample rate, 8000 digits are added each second. With T1 carriers, an alternating 1/0 frame-synchronizing pat- tern is used. To acquire frame synchronization, the digital terminal in the receiver searches through the incoming data until it finds the framing bit pattern. This encompasses testing a bit, counting off 193 more bits, and then testing again for the opposite logic condition. This process continues until a repetitive alternating 1/0 pattern is found. Initial frame synchro- nization depends on the total frame time, the number of bits per frame, and the period of each bit. Searching through all possible bit positions requires N tests, where N is the num- ber of bit positions in the frame. On average, the receiving terminal dwells at a false fram- ing position for two frame periods during a search; therefore, the maximum average syn- chronization time is synchronization time = 2NT = 2N2 tb (1) where N number of bits per frame T frame period of Ntb tb bit time For the T1 carrier, N 193, T 125 μs, and tb 0.648 μs; therefore, a maximum of 74,498 bits must be tested, and the maximum average synchronization time is 48.25 ms. 8-2 Robbed-Digit Framing When a short frame is used, added-digit framing is inefficient. This occurs with single-channel PCM systems. An alternative solution is to replace the least significant bit of every nth frame with a framing bit. This process is called robbed-digit framing. The parameter n is chosen as a compromise between reframe time and signal impairment. For n 10, the SQR is impaired by only 1 dB. Robbed-digit framing does not interrupt transmission but instead periodically re- places information bits with forced data errors to maintain frame synchronization. 8-3 Added-Channel Framing Essentially, added-channel framing is the same as added-digit framing except that digits are added in groups or words instead of as individual bits. The European time-division multi- plexing scheme previously discussed uses added-channel framing. One of the 32 time slots in each frame is dedicated to a unique synchronizing bit sequence. The average number of bits to acquire frame synchronization using added-channel framing is (2) where N number of bits per frame K number of bits in the synchronizing word For the European E1 32-channel system, N 256 and K 8. Therefore, the average number of bits needed to acquire frame synchronization is 128.5. At 2.048 Mbps, the syn- chronization time is approximately 62.7 μs. N 2 212K 12 349
- 355. FIGURE 20 Interleaving: (a) bit; (b) word 8-4 Statistical Framing With statistical framing, it is not necessary to either rob or add digits. With the gray code, the second bit is a logic 1 in the central half of the code range and a logic 0 at the extremes. Therefore, a signal that has a centrally peaked amplitude distribution generates a high prob- ability of a logic 1 in the second digit. Hence, the second digit of a given channel can be used for the framing bit. 8-5 Unique-Line Code Framing With unique-line code framing, some property of the framing bit is different from the data bits. The framing bit is made either higher or lower in amplitude or with a different time du- ration. The earliest PCM-TDM systems used unique-line code framing. D1 channel banks used framing pulses that were twice the amplitude of normal data bits. With unique-line code framing, either added-digit or added-word framing can be used, or specified data bits can be used to simultaneously convey information and carry synchronizing signals. The ad- vantage of unique-line code framing is that synchronization is immediate and automatic. The disadvantage is the additional processing requirements necessary to generate and rec- ognize the unique bit. 9 BIT VERSUS WORD INTERLEAVING When time-division multiplexing two or more PCM systems, it is necessary to interleave the transmissions from the various terminals in the time domain. Figure 20 shows two methods of interleaving PCM transmissions: bit interleaving and word interleaving. T1 carrier systems use word interleaving; eight-bit samples from each channel are in- terleaved into a single 24-channel TDM frame. Higher-speed TDM systems and delta mod- Digital T-Carriers and Multiplexing 350
- 356. Digital T-Carriers and Multiplexing ulation systems use bit interleaving. The decision as to which type of interleaving to use is usually determined by the nature of the signals to be multiplexed. 10 STATISTICAL TIME-DIVISION MULTIPLEXING Digital transmissions over a synchronous TDM system often contain an abundance of time slots within each frame that contain no information (i.e., at any given instant, several of the channels may be idle). For example, TDM is commonly used to link remote data terminals or PCs to a common server or mainframe computer. A majority of the time, however, there are no data being transferred in either direction, even if all the terminals are active. The same is true for PCM-TDM systems carrying digital-encoded voice-grade telephone con- versations. Normal telephone conversations generally involve information being trans- ferred in only one direction at a time with significant pauses embedded in typical speech patterns. Consequently, there is a lot of time wasted within each TDM frame. There is an efficient alternative to synchronous TDM called statistical time-division multiplexing. Sta- tistical time division multiplexing is generally not used for carrying standard telephone cir- cuits but are used more often for the transmission of data when they are called asynchronous TDM, intelligent TDM, or simply stat muxs. A statistical TDM multiplexer exploits the natural breaks in transmissions by dynami- cally allocating time slots on a demand basis. Just as with the multiplexer in a synchronous TDM system, a statistical multiplexer has a finite number of low-speed data input lines with one high-speed multiplexed data output line, and each input line has its own digital encoder and buffer.With the statistical multiplexer, there are ninput lines but only ktime slots available within the TDM frame (where k n). The multiplexer scans the input buffers, collecting data untilaframeisfilled,atwhichtimetheframeistransmitted.Onthereceiveend,thesameholds true, as there are more output lines than time slots within the TDM frame. The demultiplexer removes the data from the time slots and distributes them to their appropriate output buffers. Statistical TDM takes advantage of the fact that the devices attached to the inputs and outputs are not all transmitting or receiving all the time and that the data rate on the multi- plexed line is lower than the combined data rates of the attached devices. In other words, statistical TDM multiplexers require a lower data rate than synchronous multiplexers need to support the same number of inputs. Alternately, a statistical TDM multiplexer operating at the same transmission rate as a synchronous TDM multiplexer can support more users. Figure 21 shows a comparison between statistical and synchronous TDM. Four data sources (A, B, C, and D) and four time slots, or epochs (t1, t2, t3, and t4). The synchronous multiplexer has an output data rate equal to four times the data rate of each of the input channels. During each sample time, data are collected from all four sources and transmit- ted regardless of whether there is any input. As the figure shows, during sample time t1, channels C and D have no input data, resulting in a transmitted TDM frame void of infor- mation in time slots C1 and D1. With a statistical multiplexer, however, the empty time slots are not transmitted. A disadvantage of the statistical format, however, is that the length of a frame varies and the positional significance of each time slot is lost. There is no way of knowing beforehand which channel’s data will be in which time slot or how many time slots are included in each frame. Because data arrive and are distributed to receive buffers un- predictably, address information is required to ensure proper delivery. This necessitates more overhead per time slot for statistical TDM because each slot must carry an address as well as data. The frame format used by a statistical TDM multiplexer has a direct impact on system performance. Obviously, it is desirable to minimize overhead to improve data throughput. Normally, a statistical TDM system will use a synchronous protocol such as HDLC. With statistical multiplexing, control bits must be included within the frame. Figure 22a shows the 351
- 357. Digital T-Carriers and Multiplexing Sample time Wasted time slots Wasted time slot Inputs 4 to 1 TDM multiplexer TDM output A XX OO OO XX B XX XX XX OO C OO XX OO XX XX XX OO OO OO XX XX OO Synchronous A1 B1 C1 D1 A2 B2 C2 D2 TDM 1 2 3, 4, etc. Frame No wasted time slots XX XX XX XX XX XX XX XX Synchronous XX denotes data present OO denotes no data present A1 B1 B2 C2 B3 A4 C4 D4 TDM 1 2 3 4 Frame D OO OO OO XX t1 t2 t3 t4 – – – – – – – – – – – – – – – – FIGURE 21 Comparison between synchronous and statistical TDM FIGURE 22 Statistical TDM frame format: (a) overall statistical TDM frame; (b) one-source per frame; (c) multiple sources per frame overall frame format for a statistical TDM multiplexer. The frame includes beginning and ending flags that indicate the beginning and end of the frame, an address field that identifies the transmitting device, a control field, a statistical TDM subframe, and a frame check se- quence field (FCS) that provides error detection. Figure 22b shows the frame when only one data source is transmitting. The trans- mitting device is identified in the address field. The data field length is variable and limited only by the maximum length of the frame. Such a scheme works well in times of light loads but rather inefficiently under heavy loads. Figure 14c shows one way to improve the effi- ciency by allowing more than one data source to be included within a single frame. With multiple sources, however, some means is necessary to specify the length of the data stream 352
- 358. from each source. Hence, the statistical frame consists of sequences of data fields labeled with an address and a bit count. There are several techniques that can be used to further im- prove efficiency. The address field can be shortened by using relative addressing where each address specifies the position of the current source relative to the previously transmitted source and the total number of sources. With relative addressing, an eight-bit address field can be replaced with a four-bit address field. Another method of refining the frame is to use a two-bit label with the length field. The binary values 01, 10, and 11 correspond to a data field of 1, 2, or 3 bytes, respectively, and no length field necessary is indicated by the code 00. A codec is a large-scale integration (LSI) chip designed for use in the telecommunications industry for private branch exchanges (PBXs), central office switches, digital handsets, voice store-and-forward systems, and digital echo suppressors. Essentially, the codec is ap- plicable for any purpose that requires the digitizing of analog signals, such as in a PCM- TDM carrier system. Codec is a generic term that refers to the coding functions performed by a device that converts analog signals to digital codes and digital codes to analog signals. Recently de- veloped codecs are called combo chips because they combine codec and filter functions in the same LSI package. The input/output filter performs the following functions: bandlimit- ing, noise rejection, antialiasing, and reconstruction of analog audio waveforms after de- coding. The codec performs the following functions: analog sampling, encoding/decoding (analog-to-digital and digital-to-analog conversions), and digital companding. A combo chip can provide the analog-to-digital and the digital-to-analog conversions and the transmit and receive filtering necessary to interface a full-duplex (four-wire) voice tele- phone circuit to the PCM highway of a TDM carrier system. Essentially, a combo chip re- places the older codec and filter chip combination. Table 6 lists several of the combo chips available and their prominent features. Features of Several Codec/Filter Combo Chips 2916 (16-Pin) 2917 (16-Pin) 2913 (20-Pin) 2914 (24-Pin) μ-law companding only Master clock, 2.048 MHz only Fixed data rate Variable data rate, 64 kbps–2.048 Mbps 78-dB dynamic range ATT D3/4 compatible Single-ended input Single-ended output Gain adjust transmit only Synchronous clocks A-law companding only Master clock, 2.048 MHz only Fixed data rate Variable data rate, 64 kbps–4.096 Mbps 78-dB dynamic range ATT D3/4 compatible Single-ended input Single-ended output Gain adjust transmit only Synchronous clocks /A-law companding Master clock, 1.536 MHz, 1.544 MHz, or 2.048 MHz Fixed data rate Variable data rate, 64 kbps–4.096 Mbps 78-dB dynamic range ATT D3/4 compatible Differential input Differential output Gain adjust transmit and receive Synchronous clocks /A-law companding Master clock, 1.536 MHz, 1.544 MHz, or 2.048 MHz Fixed data rate Variable data rate, 64 kbps–4.096 Mbps 78-dB dynamic range ATT D3/4 compatible Differential input Differential output Gain adjust transmit and receive Synchronous clocks Asynchronous clocks Analog loopback Signaling Digital T-Carriers and Multiplexing 11 CODECS AND COMBO CHIPS 11-1 Codec 11-2 Combo Chips able 6 T 353
- 359. Digital T-Carriers and Multiplexing 11-2-1 General operation. The following major functions are provided by a combo chip: 1. Bandpass filtering of the analog signals prior to encoding and after decoding 2. Encoding and decoding of voice and call progress signals 3. Encoding and decoding of signaling and supervision information 4. Digital companding Figure 23a shows the block diagram of a typical combo chip. Figure 23b shows the frequency response curve for the transmit bandpass filter, and Figure 23c shows the fre- quency response for the receive low-pass filter. 11-2-2 Fixed-data-rate mode. In the fixed-data-rate mode, the master transmit and receive clocks on a combo chip (CLKX and CLKR) perform the following functions: 1. Provide the master clock for the on-board switched capacitor filter 2. Provide the clock for the analog-to-digital and digital-to-analog converters 3. Determine the input and output data rates between the codec and the PCM highway Therefore, in the fixed-data-rate mode, the transmit and receive data rates must be ei- ther 1.536 Mbps, 1.544 Mbps, or 2.048 Mbps—the same as the master clock rate. Transmit and receive frame synchronizing pulses (FSX and FSR) are 8-kHz in- puts that set the transmit and receive sampling rates and distinguish between signaling and nonsignaling frames. is a time-slot strobe buffer enable output that is used to gate the PCM word onto the PCM highway when an external buffer is used to drive the line. is also used as an external gating pulse for a time-division multiplexer (see Figure 24a). Data are transmitted to the PCM highway from DX on the first eight positive transi- tions of CLKX following the rising edge of FSX. On the receive channel, data are received from the PCM highway from DR on the first eight falling edges of CLKR after the occur- rence of FSR. Therefore, the occurrence of FSX and FSR must be synchronized between codecs in a multiple-channel system to ensure that only one codec is transmitting to or re- ceiving from the PCM highway at any given time. Figures 24a and b show the block diagram and timing sequence for a single-chan- nel PCM system using a combo chip in the fixed-data-rate mode and operating with a mas- ter clock frequency of 1.536 MHz. In the fixed-data-rate mode, data are input and output for a single channel in short bursts. (This mode of operation is sometimes called the burst mode.) With only a single channel, the PCM highway is active only 1/24 of the total frame time.Additional channels can be added to the system provided that their transmissions are synchronized so that they do not occur at the same time as transmissions from any other channel. From Figure 24, the following observations can be made: 1. The input and output bit rates from the codec are equal to the master clock fre- quency, 1.536 Mbps. 2. The codec inputs and outputs 64,000 PCM bits per second. 3. The data output (DX) and data input (DR) are enabled only 1/24 of the total frame time (125 μs). To add channels to the system shown in Figure 24, the occurrence of the FSX, FSR, and signals for each additional channel must be synchronized so that they follow a timely sequence and do not allow more than one codec to transmit or receive at the same TSX TSX TSX 354
- 361. Digital T-Carriers and Multiplexing +0.15 dB 300 Hz +0.15 dB 3000 Hz Expanded scale Gain relative to gain at 1 kHz – dB –0.15 dB 3000 Hz –0.15 dB 300 Hz Typical filter transfer function –0.35 dB 3300 Hz –1.0 dB 3400 Hz –14 dB 4000 Hz –32 dB 4600 Hz +0.03 dB 3300 Hz –0.10 dB 3400 Hz 0 –1 –10 –20 –30 –40 –50 –60 10 k 1 k 3500 200 100 50 10 –60 –50 –40 –30 –20 –10 –1 0 0 0 0 dB 1000 Hz –0.125 dB 200 Hz –1.8 dB 200 Hz –30 dB 16.67 Hz –23 dB 60 Hz –25 dB 50 Hz Typical filter transfer function (b) Frequency – Hz –14 dB 4000 Hz –30 dB 4800 Hz –0.15 dB 300 Hz –0.5 dB 200 Hz +0.125 dB 200 Hz +2 +1 –1 –10 –20 –30 –40 –50 100 200 3300 1k Frequency – Hz (c) 10k 0 0 +2 +1 –1 –10 –20 –30 –40 –50 0 0 +0.15 dB 300 Hz +0.15 dB 3000 Hz +0.03 dB 3300 Hz –0.10 dB 3400 Hz –0.15 dB 3000 Hz –0.35 dB 3300 Hz –1.0 dB 3400 Hz 0 dB 1000 Hz Expanded scale Gain relative to gain at 1 kHz – dB FIGURE 23 (Continued) Combo chip: (b) transmit BPF response curve; (c) receive LPF response curve 356
- 362. Digital T-Carriers and Multiplexing FIGURE 24 Single-channel PCM system using a combo chip in the fixed-data-rate mode: (a) block diagram; (Continued) time. Figures 25a and b show the block diagram and timing sequence for a 24-channel PCM-TDM system operating with a master clock frequency of 1.536 MHz. 11-2-3 Variable-data-rate mode. The variable-data-rate mode allows for a flexible data input and output clock frequency. It provides the ability to vary the frequency of the transmit and receive bit clocks. In the variable-data-rate mode, a master clock frequency of 1.536 MHz, 1.544 MHz, or 2.048 MHz is still required for proper operation of the onboard bandpass filters and the analog-to-digital and digital-to-analog converters. However, in the variable-data-rate mode, DCLKR and DCLKX become the data clocks for the receive and transmit PCM highways, respectively. When FSX is high, data are transmitted onto the PCM highway on the next eight consecutive positive transitions of DCLKX. Similarly, while FSR is high, data from the PCM highway are clocked into the codec on the next eight consecutive negative transitions of DCLKR. This mode of operation is sometimes called the shift register mode. On the transmit channel, the last transmitted PCM word is repeated in all remaining time slots in the 125-μs frame as long as DCLKX is pulsed and FSX is held active high. 357
- 366. Digital T-Carriers and Multiplexing This feature allows the PCM word to be transmitted to the PCM highway more than once per frame. Signaling is not allowed in the variable-data-rate mode because this mode pro- vides no means to specify a signaling frame. Figures 26a and b shows the block diagram and timing sequence for a two-channel PCM-TDMsystemusingacombochipinthevariable-data-ratemodewithamasterclockfre- quency of 1.536 MHz, a sample rate of 8 kHz, and a transmit and receive data rate of 128 kbps. With a sample rate of 8 kHz, the frame time is 125 μs. Therefore, one eight-bit PCM word from each channel is transmitted and/or received during each 125-μs frame. For 16 bits to occur in 125 μs, a 128-kHz transmit and receive data clock is required: or 8 bits channel 2 channels frame 8000 frames second 128 kbps bit rate 1 tb 1 7.8125 μs 128 kbps tb 1 channel 8 bits 1 frame 2 channels 125 μs frame 125 μs 16 bits 7.8125 μs bit FIGURE 26 Two-channel PCM-TDM system using a combo chip in the variable-data-rate mode with a master clock frequency of 1.536 MHz: (a) block diagram; (Continued) 361
- 367. Digital T-Carriers and Multiplexing The transmit and receive enable signals (FSX and FSR) for each codec are active for one-half of the total frame time. Consequently, 8-kHz, 50% duty cycle transmit and receive data enable signals (FSX and FXR) are fed directly to one codec and fed to the other codec 180° out of phase (inverted), thereby enabling only one codec at a time. To expand to a four-channel system, simply increase the transmit and receive data clock rates to 256 kHz and change the enable signals to 8-kHz, 25% duty cycle pulses. 11-2-4 Supervisory signaling. With a combo chip, supervisory signaling can be used only in the fixed-data-rate mode. A transmit signaling frame is identified by making the FSX and FSR pulses twice their normal width. During a transmit signaling frame, the signal present on input SIGX is substituted into the least significant bit position (b1) of the encoded PCM word. At the receive end, the signaling bit is extracted from the PCM word prior to decoding and placed on output SIGR until updated by reception of another signaling frame. Asynchronous operation occurs when the master transmit and receive clocks are de- rived from separate independent sources. A combo chip can be operated in either the syn- chronous or the asynchronous mode using separate digital-to-analog converters and volt- age references in the transmit and receive channels, which allows them to be operated FIGURE 26 (Continued) (b) timing diagram 362
- 368. Digital T-Carriers and Multiplexing Channel 1 Station 1 Channel 2 Channel 3 B = 10 kHz B = 10 kHz B = 10 kHz Channel 4 – 104 Channel 105 Channel 106 Channel 107 535 Station 2 Station 3 Station 106 Station 107 B = 10 kHz B = 10 kHz B = 10 kHz 545 555 565 1575 1585 1595 1605 FIGURE 27 Frequency-division multiplexing of the commercial AM broadcast band completely independent of each other. With either synchronous or asynchronous operation, the master clock, data clock, and time-slot strobe must be synchronized at the beginning of each frame. In the variable-data-rate mode, CLKX and DCLKX must be synchronized once per frame but may be different frequencies. 12 FREQUENCY-DIVISION MULTIPLEXING With frequency-division multiplexing (FDM), multiple sources that originally occupied the same frequency spectrum are each converted to a different frequency band and transmitted si- multaneously over a single transmission medium, which can be a physical cable or the Earth’s atmosphere (i.e., wireless). Thus, many relatively narrow-bandwidth channels can be trans- mitted over a single wide-bandwidth transmission system without interfering with each other. FDM is used for combining many relatively narrowband sources into a single wideband chan- nel, such as in public telephone systems. Essentially, FDM is taking a given bandwidth and subdividing it into narrower segments with each segment carrying different information. FDMisananalogmultiplexingscheme;theinformationenteringanFDMsystemmust be analog, and it remains analog throughout transmission. If the original source information is digital, it must be converted to analog before being frequency-division multiplexed. A familiar example of FDM is the commercial AM broadcast band, which occupies a frequency spectrum from 535 kHz to 1605 kHz. Each broadcast station carries an infor- mation signal (voice and music) that occupies a bandwidth between 0 Hz and 5 kHz. If the information from each station were transmitted with the original frequency spectrum, it would be impossible to differentiate or separate one station’s transmissions from another. Instead, each station amplitude modulates a different carrier frequency and produces a 10- kHz signal. Because the carrier frequencies of adjacent stations are separated by 10 kHz, the total commercialAM broadcast band is divided into 107 10-kHz frequency slots stacked next to each other in the frequency domain. To receive a particular station, a receiver is sim- ply tuned to the frequency band associated with that station’s transmissions. Figure 27 shows how commercial AM broadcast station signals are frequency-division multiplexed and transmitted over a common transmission medium (Earth’s atmosphere). With FDM, each narrowband channel is converted to a different location in the total frequency spectrum. The channels are stacked on top of one another in the frequency domain. Figure 28a shows a simple FDM system where four 5-kHz channels are fre- quency-division multiplexed into a single 20-kHz combined channel. As the figure shows, 363
- 369. Digital T-Carriers and Multiplexing Channel 1 0–5 kHz BPF 100 kHz oscillator Balanced modulator DSBSC 95–105 kHz DSBSC 100–110 kHz Balanced modulator Linear summer FDM Output 100–120 kHz 105 kHz oscillator x Channel 2 0–5 kHz BPF x DSBSC 105–115 kHz Balanced modulator 110 kHz oscillator Channel 3 0–5 kHz BPF x DSBSC 110–120 kHz SSBSC 100–105 kHz SSBSC 105–110 kHz SSBSC 110–115 kHz SSBSC 115–120 kHz Balanced modulator 115 kHz oscillator Channel 4 0–5 kHz BPF (a) x FIGURE 28 Frequency-division multiplexing: (a) block diagram; (b) frequency spectrum Channel 1 source information 100 kHz 105 kHz 110 kHz 115 kHz 120 kHz Channel 2 source information Channel 3 source information Channel 4 source information Bandwidth = 20 kHz (b) 364
- 370. Digital T-Carriers and Multiplexing channel 1 signals amplitude modulate a 100-kHz carrier in a balanced modulator, which in- herently suppresses the 100-kHz carrier. The output of the balanced modulator is a double- sideband suppressed-carrier waveform with a bandwidth of 10 kHz. The double sideband waveform passes through a bandpass filter (BPF) where it is converted to a single sideband signal. For this example, the lower sideband is blocked; thus, the output of the BPF occu- pies the frequency band between 100 kHz and 105 kHz (a bandwidth of 5 kHz). Channel 2 signals amplitude modulate a 105-kHz carrier in a balanced modulator, again producing a double sideband signal that is converted to single sideband by passing it through a bandpass filter tuned to pass only the upper sideband. Thus, the output from the BPF occupies a frequency band between 105 kHz and 110 kHz. The same process is used to convert signals from channels 3 and 4 to the frequency bands 110 kHz to 115 kHz and 115 kHz to 120 kHz, respectively. The combined frequency spectrum produced by com- bining the outputs from the four bandpass filters is shown in Figure 28b. As the figure shows, the total combined bandwidth is equal to 20 kHz, and each channel occupies a dif- ferent 5-kHz portion of the total 20-kHz bandwidth. There are many other applications for FDM, such as commercial FM and television broadcasting, high-volume telephone and data communications systems, and cable televi- sion and data distribution networks. Within any of the commercial broadcast frequency bands, each station’s transmissions are independent of all the other stations’ transmissions. Consequently, the multiplexing (stacking) process is accomplished without synchroniza- tion between stations. With a high-volume telephone communications system, many voice- band telephone channels may originate from a common source and terminate in a common destination. The source and destination terminal equipment is most likely a high-capacity electronic switching system (ESS). Because of the possibility of a large number of narrow- band channels originating and terminating at the same location, all multiplexing and de- multiplexing operations must be synchronized. 13 ATT’S FDM HIERARCHY Although ATT is no longer the only long-distance common carrier in the United States, it still provides the vast majority of the long-distance services and, if for no other reason than its overwhelming size, has essentially become the standards organization for the tele- phone industry in North America. ATT’s nationwide communications network is subdivided into two classifications: short haul (short distance) and long haul (long distance). The T1 carrier explained earlier in this chapter is an example of a short-haul communications system. Figure 29 shows ATT’s long-haul FDM hierarchy. Only the transmit terminal is shown, although a complete set of inverse functions must be performed at the re- ceiving terminal. As the figure shows, voice channels are combined to form groups, groups are combined to form supergroups, and supergroups are combined to form mastergroups. 13-1 Message Channel The message channel is the basic building block of the FDM hierarchy. The basic message channel was originally intended for the analog voice transmission, although it now includes any transmissions that utilize voice-band frequencies (0 kHz to 4 kHz), such as data trans- mission using voice-band data modems. The basic voice-band (VB) circuit is called a basic 3002 channel and is actually bandlimited to approximately a 300-Hz to 3000-Hz frequency band, although for practical design considerations it is considered a 4-kHz channel. The basic 3002 channel can be subdivided and frequency-division multiplexed into 24 narrower-band 3001 (telegraph) channels. 365
- 371. Digital T-Carriers and Multiplexing Voice-band data modem Channel 1 Voiceband telephone channels 2-12 Other channel banks 256 kbps modem Group 2 Group 3 Group 4 Group 5 To Channel bank (12 VB channels) Group bank (60 VB channels) Supergroup bank (600 VB channels) Master- group bank (1800 VB channels) To radio, coaxial cable, optical fiber, or satellite transmitter Group 1 SG 11 SG 13-17 SG 12 SG D18 SG D25 MG 2 From other supergroup banks From other group banks MG 3 SG D26 SG D27 SG D28 MG 1 FIGURE 29 American Telephone Telegraph Company’s FDM hierarchy 13-2 Basic Group A group is the next higher level in the FDM hierarchy above the basic message channel and, consequently, is the first multiplexing step for combining message channels. A basic group consists of 12 voice-band message channels multiplexed together by stacking them next to each other in the frequency domain. Twelve 4-kHz voice-band channels occupy a combined bandwidth of 48 kHz (4 12). The 12-channel modulating block is called an A-type (analog) channel bank. The 12-channel group output from an A-type channel bank is the standard building block for most long-haul broadband telecommunications systems. 13-3 Basic Supergroup The next higher level in the FDM hierarchy shown in Figure 29 is the supergroup, which is formed by frequency-division multiplexing five groups containing 12 channels each for a combined bandwidth of 240 kHz (5 groups 48 kHz/group or 5 groups 12 channels/ group 4 kHz/channel). 366
- 372. Digital T-Carriers and Multiplexing 13-4 Basic Mastergroup The next highest level of multiplexing shown in Figure 29 is the mastergroup, which is formed by frequency-division multiplexing 10 supergroups together for a combined capac- ity of 600 voice-band message channels occupying a bandwidth of 2.4 MHz (600 channels 4 kHz/channel or 5 groups 12/channels/group 10 groups/supergroup). Typically, three mastergroups are frequency-division multiplexed together and placed on a single mi- crowave or satellite radio channel. The capacity is 1800 VB channels (3 mastergroups 600 channels/mastergroup) utilizing a combined bandwidth of 7.2 MHz. 13-5 Larger Groupings Mastergroups can be further multiplexed in mastergroup banks to form jumbogroups (3600 VB channels), multijumbogroups (7200 VB channels), and superjumbogroups (10,800 VB channels). 14 COMPOSITE BASEBAND SIGNAL Baseband describes the modulating signal (intelligence) in a communications system.A sin- gle message channel is baseband.A group, supergroup, or mastergroup is also baseband. The composite baseband signal is the total intelligence signal prior to modulation of the final car- rier. In Figure 29, the output of a channel bank is baseband. Also, the output of a group or supergroup bank is baseband. The final output of the FDM multiplexer is the composite (to- tal) baseband. The formation of the composite baseband signal can include channel, group, supergroup, and mastergroup banks, depending on the capacity of the system. 14-1 Formation of Groups and Supergroups Figure 30 shows how a group is formed with an A-type channel bank. Each voice-band channel is bandlimited with an antialiasing filter prior to modulating the channel carrier. FDM uses single-sideband suppressed-carrier (SSBSC) modulation. The combination of the balanced modulator and the bandpass filter makes up the SSBSC modulator. A balanced modulator is a double-sideband suppressed-carrier modulator, and the band- pass filter is tuned to the difference between the carrier and the input voice-band fre- quencies (LSB). The ideal input frequency range for a single voice-band channel is 0 kHz to 4 kHz. The carrier frequencies for the channel banks are determined from the fol- lowing expression: fc = 112 4n kHz (3) where n is the channel number. Table 7 lists the carrier frequencies for channels 1 through 12. Therefore, for channel 1, a 0-kHz to 4-kHz band of frequencies modulates a 108-kHz carrier. Mathematically, the output of a channel bandpass filter is fout = (fc 4 kHz) to fc (4) where fc channel carrier frequency (112 4n kHz) and each voice-band channel has a 4-kHz bandwidth. For channel 1, fout 108 kHz 4 kHz 104 kHz to 108 kHz For channel 2, fout 104 kHz 4 kHz 100 kHz to 104 kHz For channel 12, fout 64 kHz 4 kHz 60 kHz to 64 kHz The outputs from the 12 A-type channel modulators are summed in the linear com- biner to produce the total group spectrum shown in Figure 30b (60 kHz to 108 kHz). Note that the total group bandwidth is equal to 48 kHz (12 channels 4 kHz). 367
- 374. Digital T-Carriers and Multiplexing Figure 31a shows how a supergroup is formed with a group bank and combining net- work. Five groups are combined to form a supergroup. The frequency spectrum for each group is 60 kHz to 108 kHz. Each group is mixed with a different group carrier frequency in a balanced modulator and then bandlimited with a bandpass filter tuned to the difference frequency band (LSB) to produce a SSBSC signal. The group carrier frequencies are de- rived from the following expression: fc 372 48n kHz where n is the group number. Table 8 lists the carrier frequencies for groups 1 through 5. For group 1, a 60-kHz to 80-kHz group signal modulates a 420-kHz group carrier fre- quency. Mathematically, the output of a group bandpass filter is fout (fc 108 kHz) to (fc 60 kHz) where fc group carrier frequency (372 48n kHz) and for a group frequency spectrum of 60 KHz to 108 KHz Group 1, fout 420 kHz (60 kHz to 108 kHz) 312 kHz to 360 kHz Group 2, fout 468 kHz (60 kHz to 108 kHz) 360 kHz to 408 kHz Group 5, fout 612 kHz (60 kHz to 108 kHz) 504 kHz to 552 kHz The outputs from the five group modulators are summed in the linear combiner to produce the total supergroup spectrum shown in Figure 31b (312 kHz to 552 kHz). Note that the total supergroup bandwidth is equal to 240 kHz (60 channels 4 kHz). 15 FORMATION OF A MASTERGROUP There are two types of mastergroups: L600 and U600 types. The L600 mastergroup is used for low-capacity microwave systems, and the U600 mastergroup may be further multi- plexed and used for higher-capacity microwave radio systems. 15-1 U600 Mastergroup Figure 32a shows how a U600 mastergroup is formed with a supergroup bank and com- bining network. Ten supergroups are combined to form a mastergroup. The frequency spec- trum for each supergroup is 312 kHz to 552 kHz. Each supergroup is mixed with a differ- ent supergroup carrier frequency in a balanced modulator. The output is then bandlimited to the difference frequency band (LSB) to form a SSBSC signal. The 10 supergroup carrier Table 8 Group Carrier Frequencies Group Carrier Frequency (kHz) 1 420 2 468 3 516 4 564 5 612 Table 7 Channel Carrier Frequencies Channel Carrier Frequency (kHz) 1 108 2 104 3 100 4 96 5 92 6 88 7 84 8 80 9 76 10 72 11 68 12 64 369
- 377. Digital T-Carriers and Multiplexing frequencies are listed in Table 9. For supergroup 13, a 312-kHz to 552-kHz supergroup band of frequencies modulates a 1116-kHz carrier frequency. Mathematically, the output from a supergroup bandpass filter is fout fc fs to fc where fc supergroup carrier frequency fs supergroup frequency spectrum (312 kHz to 552 kHz) For supergroup 13, fout 1116 kHz (312 kHz to 552 kHz) 564 kHz to 804 kHz For supergroup 14, fout 1364 kHz (312 kHz to 552 kHz) 812 kHz to 1052 kHz For supergroup D28, fout 3396 kHz (312 kHz to 552 kHz) 2844 kHz to 3084 kHz Theoutputsfromthe10supergroupmodulatorsaresummedinthelinearsummertopro- duce the total mastergroup spectrum shown in Figure 32b (564 kHz to 3084 kHz). Note that between any two adjacent supergroups, there is a void band of frequencies that is not included within any supergroup band.These voids are called guard bands.The guard bands are necessary because the demultiplexing process is accomplished through filtering and down-converting. Without the guard bands, it would be difficult to separate one supergroup from an adjacent su- pergroup. The guard bands reduce the quality factor (Q) required to perform the necessary fil- tering.Theguardbandis8kHzbetweenallsupergroupsexcept18andD25,whereitis56kHz. Consequently, the bandwidth of a U600 mastergroup is 2520 kHz (564 kHz to 3084 kHz), whichisgreaterthanisnecessarytostack600voice-bandchannels(6004kHz2400kHz). Guard bands were not necessary between adjacent groups because the group fre- quencies are sufficiently low, and it is relatively easy to build bandpass filters to separate one group from another. In the channel bank, the antialiasing filter at the channel input passes a 0.3-kHz to 3-kHz band. The separation between adjacent channel carrier frequencies is 4 kHz. Therefore, there is a 1300-Hz guard band between adjacent channels. This is shown in Figure 33. 15-2 L600 Mastergroup With an L600 mastergroup, 10 supergroups are combined as with the U600 mastergroup, except that the supergroup carrier frequencies are lower. Table 10 lists the supergroup carrier frequencies for an L600 mastergroup. With an L600 mastergroup, the composite baseband spectrum occupies a lower-frequency band than the U-type mastergroup (Figure 34). An L600 mastergroup is not further multiplexed. Therefore, the maximum channel capacity for a microwave or coaxial cable system using a single L600 master- group is 600 voice-band channels. Table 9 Supergroup Carrier Frequencies for a U600 Mastergroup Carrier Frequency Supergroup (kHz) 13 1116 14 1364 15 1612 16 1860 17 2108 18 2356 D25 2652 D26 2900 D27 3148 D28 3396 372
- 378. Digital T-Carriers and Multiplexing FIGURE 33 Channel guard bands Table 10 Supergroup Carrier Frequencies for a L600 Mastergroup Carrier Frequency Supergroup (kHz) 1 612 2 Direct 3 1116 4 1364 5 1612 6 1860 7 2108 8 2356 9 2724 10 3100 FIGURE 34 L600 mastergroup 15-3 Formation of a Radio Channel A radio channel comprises either a single L600 mastergroup or up to three U600 master- groups (1800 voice-band channels). Figure 35a shows how an 1800-channel composite FDM baseband signal is formed for transmission over a single microwave radio channel. Mastergroup 1 is transmitted directly as is, while mastergroups 2 and 3 undergo an addi- tional multiplexing step. The three mastergroups are summed in a mastergroup combining network to produce the output spectrum shown in Figure 35b. Note the 80-kHz guard band between adjacent mastergroups. The system shown in Figure 35 can be increased from 1800 voice-band channels to 1860 by adding an additional supergroup (supergroup 12) directly to mastergroup 1. The additional 312-kHz to 552-kHz supergroup extends the composite output spectrum from 312 kHz to 8284 kHz. 373
- 380. Digital T-Carriers and Multiplexing 16 WAVELENGTH-DIVISION MULTIPLEXING During the last two decades of the 20th century, the telecommunications industry witnessed an unprecedented growth in data traffic and the need for computer networking. The possi- bility of using wavelength-division multiplexing (WDM) as a networking mechanism for telecommunications routing, switching, and selection based on wavelength begins a new era in optical communications. WDM promises to vastly increase the bandwidth capacity of optical transmission me- dia. The basic principle behind WDM involves the transmission of multiple digital signals using several wavelengths without their interfering with one another. Digital transmission equipment currently being deployed utilizes optical fibers to carry only one digital signal per fiber per propagation direction. WDM is a technology that enables many optical signals to be transmitted simultaneously by a single fiber cable. WDM is sometimes referred to as simply wave-division multiplexing. Since wave- length and frequency are closely related, WDM is similar to frequency-division multiplex- ing (FDM) in that the idea is to send information signals that originally occupied the same band of frequencies through the same fiber at the same time without their interfering with each other. This is accomplished by modulating injection laser diodes that are transmitting highly concentrated light waves at different wavelengths (i.e., at different optical frequen- cies). Therefore, WDM is coupling light at two or more discrete wavelengths into and out of an optical fiber. Each wavelength is capable of carrying vast amounts of information in either analog or digital form, and the information can already be time- or frequency- division multiplexed. Although the information used with lasers is almost always time- division multiplexed digital signals, the wavelength separation used with WDM is analo- gous to analog radio channels operating at different carrier frequencies. However, the carrier with WDM is in essence a wavelength rather than a frequency. 16-1 WDM versus FDM The basic principle of WDM is essentially the same as FDM, where several signals are trans- mittedusingdifferentcarriers,occupyingnonoverlappingbandsofafrequencyorwavelength spectrum. In the case ofWDM, the wavelength spectrum used is in the region of 1300 or 1500 nm, which are the two wavelength bands at which optical fibers have the least amount of sig- nal loss. In the past, each window transmitted a signal digital signal. With the advance of op- tical components, each transmitting window can be used to propagate several optical signals, each occupying a small fraction of the total wavelength window. The number of optical sig- nals multiplexed with a window is limited only by the precision of the components used. Cur- rent technology allows over 100 optical channels to be multiplexed into a single optical fiber. Although FDM and WDM share similar principles, they are not the same. The most obvious difference is that optical frequencies (in THz) are much higher than radio frequen- cies (in MHz and GHz). Probably the most significant difference, however, is in the way the two signals propagate through their respective transmission media. With FDM, signals propagate at the same time and through the same medium and follow the same transmission path. The basic principle of WDM, however, is somewhat different. Different wavelengths in a light pulse travel through an optical fiber at different speeds (e.g., blue light propagates slower than red light). In standard optical fiber communications systems, as the light prop- agates down the cable, wavelength dispersion causes the light waves to spread out and dis- tribute their energy over a longer period of time. Thus, in standard optical fiber systems, wavelength dispersion creates problems, imposing limitations on the system’s performance. With WDM, however, wavelength dispersion is the essence of how the system operates. With WDM, information signals from multiple sources modulate lasers operating at different wavelengths. Hence, the signals enter the fiber at the same time and travel through the same medium. However, they do not take the same path down the fiber. Since each 375
- 381. Digital T-Carriers and Multiplexing Channel 4 Channel 3 Channel 2 Channel 1 Time (seconds) (a) Fiber cable (b) Channel 4 bandwidth Channel 3 bandwidth Channel 2 bandwidth Channel 1 bandwidth f4 f3 f2 f1 Fiber cladding Fiber core Wavelength 2 out Wavelength 1 out Wavelength 1 in Wavelength 2 in FIGURE 36 (a) Frequency-division multiplexing; (b) wave-length-division multiplexing wavelength takes a different transmission path, they each arrive at the receive end at slightly different times. The result is a series of rainbows made of different colors (wavelengths) each about 20 billionths of a second long, simultaneously propagating down the cable. Figure 36 illustrates the basic principles of FDM and WDM signals propagating through their respective transmission media. As shown in Figure 36a, FDM channels all propagate at the same time and over the same transmission medium and take the same trans- mission path, but they occupy different bandwidths. In Figure 36b, it can be seen that with WDM, each channel propagates down the same transmission medium at the same time, but each channel occupies a different bandwidth (wavelength), and each wavelength takes a dif- ferent transmission path. 16-2 Dense-Wave-Division Multiplexing, Wavelengths, and Wavelength Channels WDM is generally accomplished at approximate wavelengths of 1550 nm (1.55 μm) with suc- cessive frequencies spaced in multiples of 100 GHz (e.g., 100 GHz, 200 GHz, 300 GHz, and so on). At 1550-nm and 100-GHz frequency separation, the wavelength separation is ap- proximately 0.8 nm. For example, three adjacent wavelengths each separated by 100 GHz 376
- 382. Digital T-Carriers and Multiplexing correspond to wavelengths of 1550.0 nm, 1549.2 nm, and 1548.4 nm. Using a multiplexing technique called dense-wave-division multiplexing (D-WDM), the spacing between adja- centfrequenciesisconsiderablyless.Unfortunately,theredoesnotseemtobeastandarddef- inition of exactly what D-WDM means. Generally, optical systems carrying multiple opti- cal signals spaced more than 200 GHz or 1.6 nm apart in the vicinity of 1550 nm are considered standard WDM. WDM systems carrying multiple optical signals in the vicinity of 1550 nm with less than 200-GHz separation are considered D-WDM. Obviously, the more wavelengths used in a WDM system, the closer they are to each other and the denser the wavelength spectrum. Light waves are comprised of many frequencies (wavelengths), and each frequency corresponds to a different color. Transmitters and receivers for optical fiber systems have been developed that transmit and receive only a specific color (i.e., a specific wavelength at a specific frequency with a fixed bandwidth). WDM is a process in which different sources of information (channels) are propagated down an optical fiber on different wave- lengths where the different wavelengths do not interfere with each other. In essence, each wavelength adds an optical lane to the transmission superhighway, and the more lanes there are, the more traffic (voice, data, video, and so on) can be carried on a single optical fiber cable. In contrast, conventional optical fiber systems have only one channel per cable, which is used to carry information over a relatively narrow bandwidth. A Bell Laboratories research team recently constructed a D-WDM transmitter using a single femtosecond, erbium-doped fiber-ring laser that can simultaneously carry 206 digitally modulated wave- lengths of color over a single optical fiber cable. Each wavelength (channel) has a bit rate of 36.7 Mbps with a channel spacing of approximately 36 GHz. Figure 37a shows the wavelength spectrum for a WDM system using six wave- lengths, each modulated with equal-bandwidth information signals. Figure 37b shows how the output wavelengths from six lasers are combined (multiplexed) and then prop- agated over a single optical cable before being separated (demultiplexed) at the receiver with wavelength selective couplers. Although it has been proven that a single, ultrafast light source can generate hundreds of individual communications channels, standard WDM communications systems are generally limited to between 2 and 16 channels. WDM enhances optical fiber performance by adding channels to existing cables. Each wavelength added corresponds to adding a different channel with its own information source and transmission bit rate. Thus, WDM can extend the information-carrying capac- ity of a fiber to hundreds of gigabits per second or higher. 16-3 Advantages and Disadvantages of WDM An obvious advantage of WDM is enhanced capacity and, with WDM, full-duplex trans- mission is also possible with a single fiber. In addition, optical communications networks use optical components that are simpler, more reliable, and often less costly than their elec- tronic counterparts. WDM has the advantage of being inherently easier to reconfigure (i.e., adding or removing channels). For example, WDM local area networks have been con- structed that allow users to access the network simply by tuning to a certain wavelength. There are also limitations to WDM. Signals cannot be placed so close in the wave- length spectrum that they interfere with each other. Their proximity depends on system de- sign parameters, such as whether optical amplification is used and what optical technique is used to combine and separate signals at different wavelengths. The International Telecommunications Union adopted a standard frequency grid for D-WDM with a spacing of 100 GHz or integer multiples of 100 GHz, which at 1550 nm corresponds to a wavelength spacing of approximately 0.8 nm. With WDM, the overall signal strength should be approximately the same for each wavelength. Signal strength is affected by fiber attenuation characteristics and the degree of amplification, both of which are wavelength dependent. Under normal conditions, the 377
- 383. Digital T-Carriers and Multiplexing Bandwidth channel 6 Bandwidth channel 5 Bandwidth channel 4 Bandwidth channel 3 Bandwidth channel 2 Bandwidth channel 1 λ6 λ6 Multiplexer (a) (b) Wavelength selective couplers Laser optical sources To laser optical detectors λ5 λ4 λ3 λ2 λ1 (λ6 – λ1) λ5 λ6 λ5 λ4 λ3 λ2 λ1 λ4 λ3 λ2 λ1 Fiber cable Demultiplexer FIGURE 37 (a) Wavelength spectrum for a WDM system using six wavelengths; (b) multiplexing and demultiplexing six lasers wavelengths chosen for a system are spaced so close to one another that attenuation differs very little among them. One difference between FDM and WDM is that WDM multiplexing is performed at extremely high optical frequencies, whereas FDM is performed at relatively low radio and baseband frequencies. Therefore, radio signals carrying FDM are not limited to propagat- ing through a contained physical transmission medium, such as an optical cable. Radio sig- nals can be propagated through virtually any transmission medium, including free space. Therefore, radio signals can be transmitted simultaneously to many destinations, whereas light waves carrying WDM are limited to a two-point circuit or a combination of many two- point circuits that can go only where the cable goes. The information capacity of a single optical cable can be increased n-fold, where n represents how many different wavelengths the fiber is propagating at the same time. Each wavelength in a WDM system is modulated by information signals from different sources. Therefore, an optical communications system using a single optical cable propagating n separate wavelengths must utilize n modulators and n demodulators. 378
- 384. Digital T-Carriers and Multiplexing 16-4 WDM Circuit Components The circuit components used with WDM are similar to those used with conventional radio- wave and metallic-wire transmission systems; however, some of the names used for WDM couplers are sometimes confusing. 16-4-1 Wavelength-division multiplexers and demultiplexers. Multiplexers or combiners mix or combine optical signals with different wavelengths in a way that allows them to all pass through a single optical fiber without interfering with one another. Demult- iplexers or splitters separate signals with different wavelengths in a manner similar to the way filters separate electrical signals of different frequencies. Wavelength demultiplexers have as many outputs as there are wavelengths, with each output (wavelength) going to a different destination. Multiplexers and demultiplexers are at the terminal ends of optical fiber communications systems. 16-4-2 Wavelength-division add/drop multiplexer/demultiplexers. Add/drop multiplexer/demultiplexers are similar to regular multiplexers and demultiplexers except they are located at intermediate points in the system. Add/drop multiplexers and demulti- plexers are devices that separate a wavelength from a fiber cable and reroute it on a differ- ent fiber going in a different direction. Once a wavelength has been removed, it can be re- placed with a new signal at the same wavelength. In essence, add/drop multiplexers and demultiplexers are used to reconfigure optical fiber cables. 16-4-3 Wavelength-division routers. WDM routers direct signals of a particu- lar wavelength to a specific destination while not separating all the wavelengths present on the cable. Thus, a router can be used to direct or redirect a particular wavelength (or wavelengths) in a different direction from that followed by the other wavelengths on the fiber. 16-4-4 Wavelength-division couplers. WDM couplers enable more efficient uti- lization of the transmission capabilities of optical fibers by permitting different wave- lengths to be combined and separated. There are three basic types of WDM couplers: diffraction grating, prism, and dichroic filter. With diffraction gratings or prisms, specific wavelengths are separated from the other optic signal by reflecting them at different angles. Once a wavelength has been separated, it can be coupled into a different fiber. A dichroic filter is a mirror with a surface that has been coated with a material that permits light of only one wavelength to pass through while reflecting all other wavelengths. Therefore, the dichroic filter can allow two wavelengths to be coupled in different optical fibers. 16-4-5 WDM and the synchronous optical network. The synchronous opti- cal network (SONET) is a multiplexing system similar to conventional time-division multiplexing except SONET was developed to be used with optical fibers. The initial SONET standard is OC-1. This level is referred to as synchronous transport level 1 (STS-1). STS-1 has a 51.84-Mbps synchronous frame structure made of 28 DS-1 sig- nals. Each DS-1 signal is equivalent to a single 24-channel T1 digital carrier system. Thus, one STS-1 system can carry 672 individual voice channels (24 28). With STS- 1, it is possible to extract or add individual DS-1 signals without completely disassem- bling the entire frame. OC-48 is the second level of SONET multiplexing. It combines 48 OC-1 systems for a total capacity of 32,256 voice channels. OC-48 has a transmission bit rate of 2.48332 Gbps (2.48332 billion bits per second). A single optical fiber can carry an OC-48 system. As many as 16 OC-48 systems can be combined using wave-division multiplexing. The light spectrum is divided into 16 different wavelengths with an OC-48 system attached to each transmitter for a combined capacity of 516,096 voice channels (16 32,256). 379
- 385. Digital T-Carriers and Multiplexing QUESTIONS 1. Define multiplexing. 2. Describe time-division multiplexing. 3. Describe the Bell System T1 carrier system. 4. What is the purpose of the signaling bit? 5. What is frame synchronization? How is it achieved in a PCM-TDM system? 6. Describe the superframe format. Why is it used? 7. What is a codec? A combo chip? 8. What is a fixed-data-rate mode? 9. What is a variable-data-rate mode? 10. What is a DSX? What is it used for? 11. Explain line coding. 12. Briefly explain unipolar and bipolar transmission. 13. Briefly explain return-to-zero and nonreturn-to-zero transmission. 14. Contrast the bandwidth considerations of return-to-zero and nonreturn-to-zero transmission. 15. Contrast the clock recovery capabilities with return-to-zero and nonreturn-to-zero transmission. 16. Contrast the error detection and decoding capabilities of return-to-zero and nonreturn-to-zero transmission. 17. What is a regenerative repeater? 18. Explain B6ZS and B3ZS. When or why would you use one rather than the other? 19. Briefly explain the following framing techniques: added-digit framing, robbed-digit framing, added-channel framing, statistical framing, and unique-line code framing. 20. Contrast bit and word interleaving. 21. Describe frequency-division multiplexing. 22. Describe a message channel. 23. Describe the formation of a group, a supergroup, and a mastergroup. 24. Define baseband and composite baseband. 25. What is a guard band? When is a guard band used? 26. Describe the basic concepts of wave-division multiplexing. 27. What is the difference between WDM and D-WDM? 28. List the advantages and disadvantages of WDM. 29. Give a brief description of the following components: wavelength-division multiplexer/ demultiplexers, wavelength-division add/drop multiplexers, and wavelength-division routers. 30. Describe the three types of wavelength-division couplers. 31. Briefly describe the SONET standard, including OC-1 and OC-48 levels. PROBLEMS 1. A PCM-TDM system multiplexes 24 voice-band channels. Each sample is encoded into seven bits, and a framing bit is added to each frame. The sampling rate is 9000 samples per second. BPRZ-AMI encoding is the line format. Determine a. Line speed in bits per second. b. Minimum Nyquist bandwidth. 2. A PCM-TDM system multiplexes 32 voice-band channels each with a bandwidth of 0 kHz to 4 kHz. Each sample is encoded with an 8-bit PCM code. UPNRZ encoding is used. Determine a. Minimum sample rate. b. Line speed in bits per second. c. Minimum Nyquist bandwidth. 380
- 386. Digital T-Carriers and Multiplexing 3. For the following bit sequence, draw the timing diagram for UPRZ, UPNRZ, BPRZ, BPNRZ, and BPRZ-AMI encoding: bit stream: 1 1 1 0 0 1 0 1 0 1 1 0 0 4. Encode the following BPRZ-AMI data stream with B6ZS and B3ZS: 000000000000 5. Calculate the 12 channel carrier frequencies for the U600 FDM system. 6. Calculate the five group carrier frequencies for the U600 FDM system. 7. A PCM-TDM system multiplexes 20 voice-band channels. Each sample is encoded into eight bits, and a framing bit is added to each frame. The sampling rate is 10,000 samples per second. BPRZ-AMI encoding is the line format. Determine a. The maximum analog input frequency. b. The line speed in bps. c. The minimum Nyquist bandwidth. 8. A PCM-TDM system multiplexes 30 voice-band channels each with a bandwidth of 0 kHz to 3 kHz. Each sample is encoded with a nine-bit PCM code. UPNRZ encoding is used. Determine a. The minimum sample rate. b. The line speed in bps. c. The minimum Nyquist bandwidth. 9. For the following bit sequence, draw the timing diagram for UPRZ, UPNRZ, BPRZ, BPNRZ, and BPRZ-AMI encoding: bit stream: 1 1 0 0 0 1 0 1 0 1 10. Encode the following BPRZ-AMI data stream with B6ZS and B3ZS: 00000000000 11. Calculate the frequency range for a single FDM channel at the output of the channel, group, su- pergroup, and mastergroup combining networks for the following assignments: CH GP SG MG 2 2 13 1 6 3 18 2 4 5 D25 2 9 4 D28 3 12. Determine the frequency that a single 1-kHz test tone will translate to at the output of the chan- nel, group, supergroup, and mastergroup combining networks for the following assignments: CH GP SG MG 4 4 13 2 6 4 16 1 1 2 17 3 11 5 D26 3 13. Calculate the frequency range at the output of the mastergroup combining network for the fol- lowing assignments: GP SG MG 3 13 2 5 D25 3 1 15 1 2 17 2 14. Calculate the frequency range at the output of the mastergroup combining network for the fol- lowing assignments: SG MG 18 2 13 3 D26 1 14 1 381
- 387. Digital T-Carriers and Multiplexing ANSWERS TO SELECTED PROBLEMS 1. a. 1.521 Mbps b. 760.5 kHz 3. 5. channel f(kHz) 1 108 2 104 3 100 4 96 5 92 6 88 7 84 8 80 9 76 10 72 11 68 12 64 7. a. 5 kHz b. 1.61 Mbps c. 805 kHz 11. CH GP SG MG 100–104 364–370 746–750 746–750 84–88 428–432 1924–1928 4320–4324 92–96 526–520 2132–2136 4112–4116 72–76 488–492 2904–2908 5940–5944 13. GP SG MG MG out (kHz) 3 13 2 5540–5598 5 25 3 6704–6752 1 15 1 1248–1296 2 17 2 4504–4552 -0000-0-00000-00-0 00- 00- 382
- 388. CHAPTER OUTLINE 1 Introduction 6 Cordless Telephones 2 The Subscriber Loop 7 Caller ID 3 Standard Telephone Set 8 Electronic Telephones 4 Basic Telephone Call Procedures 9 Paging Systems 5 Call Progress Tones and Signals OBJECTIVES ■ Define communications and telecommunications ■ Define and describe subscriber loop ■ Describe the operation and basic functions of a standard telephone set ■ Explain the relationship among telephone sets, local loops, and central office switching machines ■ Describe the block diagram of a telephone set ■ Explain the function and basic operation of the following telephone set components: ringer circuit, on/off-hook circuit, equalizer circuit, speaker, microphone, hybrid network, and dialing circuit ■ Describe basic telephone call procedures ■ Define call progress tones and signals ■ Describe the following terms: dial tone, dual-tone multifrequency, multifrequency, dial pulses, station busy, equip- ment busy, ringing, ring back, and receiver on/off hook ■ Describe the basic operation of a cordless telephone ■ Define and explain the basic format of caller ID ■ Describe the operation of electronic telephones ■ Describe the basic principles of paging systems Telephone Instruments and Signals From Chapter 8 of Advanced Electronic Communications Systems, Sixth Edition. Wayne Tomasi. Copyright © 2004 by Pearson Education, Inc. Published by Pearson Prentice Hall. All rights reserved. 383
- 389. 1 INTRODUCTION Communications is the process of conveying information from one place to another. Communications requires a source of information, a transmitter, a receiver, a destina- tion, and some form of transmission medium (connecting path) between the transmitter and the receiver. The transmission path may be quite short, as when two people are talk- ing face to face with each other or when a computer is outputting information to a printer located in the same room. Telecommunications is long-distance communications (from the Greek word tele meaning “distant” or “afar”). Although the word “long” is an arbitrary term, it generally indicates that communications is taking place between a transmitter and a receiver that are too far apart to communicate effectively using only sound waves. Although often taken for granted, the telephone is one of the most remarkable de- vices ever invented. To talk to someone, you simply pick up the phone and dial a few dig- its, and you are almost instantly connected with them. The telephone is one of the simplest devices ever developed, and the telephone connection has not changed in nearly a century. Therefore, a telephone manufactured in the 1920s will still work with today’s intricate telephone system. Although telephone systems were originally developed for conveying human speech information (voice), they are now also used extensively to transport data. This is accomplished using modems that operate within the same frequency band as human voice. Anyone who uses a telephone or a data modem on a telephone circuit is part of a global communications network called the public telephone network (PTN). Because the PTN interconnects subscribers through one or more switches, it is sometimes called the public switched telephone network (PSTN). The PTN is comprised of sev- eral very large corporations and hundreds of smaller independent companies jointly re- ferred to as Telco. The telephone system as we know it today began as an unlikely collaboration of two men with widely disparate personalities: Alexander Graham Bell and Thomas A. Watson. Bell, born in 1847 in Edinburgh, Scotland, emigrated to Ontario, Canada, in 1870, where he lived for only six months before moving to Boston, Massachusetts. Watson was born in a livery stable owned by his father in Salem, Massachusetts. The two met characteristically in 1874 and invented the telephone in 1876. On March 10, 1876, one week after his patent was allowed, Bell first succeeded in transmitting speech in his lab at 5 Exeter Place in Boston. At the time, Bell was 29 years old and Watson only 22. Bell’s patent, number 174,465, has been called the most valuable ever issued. The telephone system developed rapidly. In 1877, there were only six telephones in the world. By 1881, 3,000 telephones were producing revenues, and in 1883, there were over 133,000 telephones in the United States alone. Bell and Watson left the telephone busi- ness in 1881, as Watson put it, “in better hands.” This proved to be a financial mistake, as the telephone company they left evolved into the telecommunications giant known offi- cially as the American Telephone and Telegraph Company (ATT). Because at one time ATT owned most of the local operating companies, it was often referred to as the Bell Telephone System and sometimes simply as “Ma Bell.” By 1982, the Bell System grew to an unbelievable $155 billion in assets ($256 billion in today’s dollars), with over one mil- lion employees and 100,000 vehicles. By comparison, in 1998, Microsoft’s assets were ap- proximately $10 billion. ATT once described the Bell System as “the world’s most complicated machine.” A telephone call could be made from any telephone in the United States to virtually any other telephone in the world using this machine.AlthoughATT officially divested the Bell System on January 1, 1983, the telecommunications industry continued to grow at an un- believable rate. Some estimate that more than 1.5 billion telephone sets are operating in the world today. Telephone Instruments and Signals 384
- 390. 2 THE SUBSCRIBER LOOP The simplest and most straightforward form of telephone service is called plain old telephone service (POTS), which involves subscribers accessing the public telephone network through a pair of wires called the local subscriber loop (or simply local loop). The local loop is the most fundamental component of a telephone circuit. A local loop is simply an unshielded twisted- pairtransmissionline(cablepair),consistingoftwoinsulatedconductorstwistedtogether.The insulating material is generally a polyethylene plastic coating, and the conductor is most likely a pair of 116- to 26-gauge copper wire. A subscriber loop is generally comprised of several lengthsofcopperwireinterconnectedatjunctionandcross-connectboxeslocatedinmanholes, back alleys, or telephone equipment rooms within large buildings and building complexes. The subscriber loop provides the means to connect a telephone set at a subscriber’s location to the closest telephone office, which is commonly called an end office, local ex- change office, or central office. Once in the central office, the subscriber loop is connected to an electronic switching system (ESS), which enables the subscriber to access the public telephone network. 3 STANDARD TELEPHONE SET The word telephone comes from the Greek words tele, meaning “from afar,” and phone, meaning “sound,” “voice,” or “voiced sound.” The standard dictionary defines a telephone as follows: An apparatus for reproducing sound, especially that of the human voice (speech), at a great distance, by means of electricity; consisting of transmitting and receiv- ing instruments connected by a line or wire which conveys the electric current. In essence, speech is sound in motion. However, sound waves are acoustic waves and have no electrical component. The basic telephone set is a simple analog transceiver de- signed with the primary purpose of converting speech or acoustical signals to electrical sig- nals. However, in recent years, new features such as multiple-line selection, hold, caller ID, and speakerphone have been incorporated into telephone sets, creating a more elaborate and complicated device. However, their primary purpose is still the same, and the basic func- tions they perform are accomplished in much the same way as they have always been. The first telephone set that combined a transmitter and receiver into a single handheld unit was introduced in 1878 and called the Butterstamp telephone. You talked into one end and then turned the instrument around and listened with the other end. In 1951, Western Electric Company introduced a telephone set that was the industry standard for nearly four decades (the rotary dial telephone used by your grandparents). This telephone set is called the Bell System 500-type telephone and is shown in Figure 1a. The 500-type telephone set replaced the earlier 302-type telephone set (the telephone with the hand-crank magneto, fixed microphone, hand-held earphone, and no dialing mechanism). Although there are very few 500-type telephone sets in use in the United States today, the basic functions and operation of modern telephones are essentially the same. In modern-day telephone sets, the rotary dial mechanism is replaced with a Touch-Tone keypad. The modern Touch-Tone tele- phone is called a 2500-type telephone set and is shown in Figure 1b. The quality of transmission over a telephone connection depends on the received vol- ume, the relative frequency response of the telephone circuit, and the degree of interference. In a typical connection, the ratio of the acoustic pressure at the transmitter input to the cor- responding pressure at the receiver depends on the following: The translation of acoustic pressure into an electrical signal The losses of the two customer local loops, the central telephone office equipment, and the cables between central telephone offices Telephone Instruments and Signals 385
- 391. 3 2 1 0 9 (a) 8 7 6 5 4 3 6 9 # 2 5 8 0 1 4 7 * 3 6 9 # 2 5 8 0 1 4 7 * (b) FIGURE 1 (a) 500-type telephone set; (b) 2500-type telephone set The translation of the electrical signal at the receiving telephone set to acoustic pres- sure at the speaker output 3-1 Functions of the Telephone Set The basic functions of a telephone set are as follows: 1. Notify the subscriber when there is an incoming call with an audible signal, such as a bell, or with a visible signal, such as a flashing light. This signal is analogous to an interrupt signal on a microprocessor, as its intent is to interrupt what you are doing. These signals are purposely made annoying enough to make people want to answer the telephone as soon as possible. 2. Provide a signal to the telephone network verifying when the incoming call has been acknowledged and answered (i.e., the receiver is lifted off hook). 3. Convert speech (acoustical) energy to electrical energy in the transmitter and vice versa in the receiver. Actually, the microphone converts the acoustical energy to mechanical energy, which is then converted to electrical energy. The speaker per- forms the opposite conversions. 4. Incorporate some method of inputting and sending destination telephone numbers (either mechanically or electrically) from the telephone set to the central office switch over the local loop. This is accomplished using either rotary dialers (pulses) or Touch-Tone pads (frequency tones). 5. Regulate the amplitude of the speech signal the calling person outputs onto the telephone line. This prevents speakers from producing signals high enough in am- plitude to interfere with other people’s conversations taking place on nearby cable pairs (crosstalk). 6. Incorporate some means of notifying the telephone office when a subscriber wishes to place an outgoing call (i.e., handset lifted off hook). Subscribers cannot dial out until they receive a dial tone from the switching machine. 7. Ensure that a small amount of the transmit signal is fed back to the speaker, enabling talkers to hear themselves speaking. This feedback signal is sometimes called sidetone or talkback. Sidetone helps prevent the speaker from talking too loudly. 8. Provide an open circuit (idle condition) to the local loop when the telephone is not in use (i.e., on hook) and a closed circuit (busy condition) to the local loop when the telephone is in use (off hook). 9. Provide a means of transmitting and receiving call progress signals between the central office switch and the subscriber, such as on and off hook, busy, ringing, dial pulses, Touch-Tone signals, and dial tone. Telephone Instruments and Signals 386
- 392. Central office switching machine -48 Vdc (ring) 2-Wire local subscriber loop (a) ground (tip) Telephone set switch hook microphone Plastic sheath Copper sleeve Copper ring Ring Tip Jack Copper tip Sleeve Plastic insulating rings (b) Cord Plug Ring wire Tip wire Sleeve wire FIGURE 2 (a) Simplified two-wire loop showing telephone set hookup to a local switching machine; (b) plug and jack configurations showing tip, ring, and sleeve 1 RJ11 Connector Jack End RJ-11 Plug End RJ-11 6 Conductor 2 3 6 6 1 5 4 6 1 FIGURE 3 RJ-11 Connector 3-2 Telephone Set, Local Loop, and Central Office Switching Machines Figure 2a shows how a telephone set is connected to a central office switching machine (lo- cal switch). As shown in the figure, a basic telephone set requires only two wires (one pair) from the telephone company to operate. Again, the pair of wires connecting a subscriber to the closest telephone office is called the local loop. One wire on the local loop is called the tip, and the other is called the ring. The names tip and ring come from the 1 ⁄4-inch-diameter two-conductor phone plugs and patch cords used at telephone company switchboards to in- terconnect and test circuits. The tip and ring for a standard plug and jack are shown in Figure 2b. When a third wire is used, it is called the sleeve. Since the 1960s, phone plugs and jacks have gradually been replaced in the home with a miniaturized plastic plug known as RJ-11 and a matching plastic receptacle (shown in Figure 3). RJ stands for registered jacks and is sometimes described as RJ-XX. RJ is a series of telephone connection interfaces (receptacle and plug) that are registered with the U.S. Federal Communications Commission (FCC). The term jack sometimes describes both the receptacle and the plug and sometimes specifies only the receptacle. RJ-11 is the Telephone Instruments and Signals 387
- 393. RJ-11 Connector Local loop On/off hook Equalizer Hybrid Speaker Microphone 2 wire 4 wire on switch off off on Ringer (bell or oscillator) Resistors, Capacitors, and Inductors Ring (-48 Vdc) Tip (ground) Dialing circuit – mechanical dialer or touch-tone keypad FIGURE 4 Functional block diagram of a standard telephone set most common telephone jack in use today and can have up to six conductors. Although an RJ-11 plug is capable of holding six wires in a 3 ⁄16-inch-by-3 ⁄16-inch body, only two wires (one pair) are necessary for a standard telephone circuit to operate. The other four wires can be used for a second telephone line and/or for some other special function. As shown in Figure 2a, the switching machine outputs 48 Vdc on the ring and con- nects the tip to ground. A dc voltage was used rather than an ac voltage for several reasons: (1) to prevent power supply hum, (2) to allow service to continue in the event of a power out- age, and (3) because people were afraid of ac. Minus 48 volts was selected to minimize elec- trolytic corrosion on the loop wires. The 48 Vdc is used for supervisory signaling and to providetalkbatteryforthemicrophoneinthetelephoneset.On-hook,off-hook,anddialpuls- ing are examples of supervisory signals and are described in a later section of this chapter. It should be noted that 48Vdc is the only voltage required for the operation of a standard tele- phone. However, most modern telephones are equipped with nonstandard (and often nonessential) features and enhancements and may require an additional source of ac power. 3-3 Block Diagram of a Telephone Set A standard telephone set is comprised of a transmitter, a receiver, an electrical network for equalization, associated circuitry to control sidetone levels and to regulate signal power, and necessary signaling circuitry. In essence, a telephone set is an apparatus that creates an exact likeness of sound waves with an electric current. Figure 4 shows the functional block diagram of a telephone set. The essential components of a telephone set are the ringer circuit, on/off hook circuit, equalizer circuit, hybrid circuit, speaker, microphone, and a dialing circuit. 3-3-1 Ringer circuit. The telephone ringer has been around since August 1, 1878, when Thomas Watson filed for the first ringer patent. The ringer circuit, which was originally an electromagnetic bell, is placed directly across the tip and ring of the local loop.The purpose of the ringer is to alert the destination party of incoming calls.The audible tone from the ringer must be loud enough to be heard from a reasonable distance and offensive enough to make a person want to answer the telephone as soon as possible. In modern telephones, the bell has beenreplacedwithanelectronicoscillatorconnectedtothespeaker.Today,ringingsignalscan be any imaginable sound, including buzzing, a beeping, a chiming, or your favorite melody. 3-3-2 On/off hook circuit. The on/off hook circuit (sometimes called a switch hook) is nothing more than a simple single-throw, double-pole (STDP) switch placed across Telephone Instruments and Signals 388
- 394. the tip and ring. The switch is mechanically connected to the telephone handset so that when the telephone is idle (on hook), the switch is open. When the telephone is in use (off hook), the switch is closed completing an electrical path through the microphone between the tip and ring of the local loop. 3-3-3 Equalizer circuit. Equalizers are combinations of passive components (re- sistors, capacitors, and so on) that are used to regulate the amplitude and frequency re- sponse of the voice signals. The equalizer helps solve an important transmission problem in telephone set design, namely, the interdependence of the transmitting and receiving effi- ciencies and the wide range of transmitter currents caused by a variety of local loop cables with different dc resistances. 3-3-4 Speaker. In essence, the speaker is the receiver for the telephone. The speaker converts electrical signals received from the local loop to acoustical signals (sound waves) that can be heard and understood by a human being. The speaker is connected to the local loop through the hybrid network. The speaker is typically enclosed in the handset of the telephone along with the microphone. 3-3-5 Microphone. For all practical purposes, the microphone is the transmitter for the telephone. The microphone converts acoustical signals in the form of sound pressure waves from the caller to electrical signals that are transmitted into the telephone network through the local subscriber loop. The microphone is also connected to the local loop through the hybrid network. Both the microphone and the speaker are transducers, as they convert one form of energy into another form of energy. A microphone converts acoustical energy first to mechanical energy and then to electrical energy, while the speaker performs the exact opposite sequence of conversions. 3-3-6 Hybrid network. The hybrid network (sometimes called a hybrid coil or duplex coil) in a telephone set is a special balanced transformer used to convert a two-wire circuit (the local loop) into a four-wire circuit (the telephone set) and vice versa, thus en- abling full duplex operation over a two-wire circuit. In essence, the hybrid network sepa- rates the transmitted signals from the received signals. Outgoing voice signals are typically in the 1-V to 2-V range, while incoming voice signals are typically half that value. Another function of the hybrid network is to allow a small portion of the transmit signal to be re- turned to the receiver in the form of a sidetone. Insufficient sidetone causes the speaker to raise his voice, making the telephone conversation seem unnatural. Too much sidetone causes the speaker to talk too softly, thereby reducing the volume that the listener receives. 3-3-7 Dialing circuit. The dialing circuit enables the subscriber to output signals representing digits, and this enables the caller to enter the destination telephone number. The dialing circuit could be a rotary dialer, which is nothing more than a switch connected to a mechanical rotating mechanism that controls the number and duration of the on/off condition of the switch. However, more than likely, the dialing circuit is either an electronic dial-pulsing circuit or a Touch-Tone keypad, which sends various combinations of tones representing the called digits. 4 BASIC TELEPHONE CALL PROCEDURES Figure 5 shows a simplified diagram illustrating how two telephone sets (subscribers) are interconnected through a central office dial switch. Each subscriber is connected to the switch through a local loop. The switch is most likely some sort of an electronic switching system (ESS machine). The local loops are terminated at the calling and called stations in telephone sets and at the central office ends to switching machines. Telephone Instruments and Signals 389
- 395. Called party's house Calling party's house Called party's 2-Wire Local Loop Calling party's 2-Wire Local Loop Local Telephone Office Switching Machine Called party's telephone set Calling party's telephone set RJ-11 Connector RJ-11 Connector R (-48 Vdc) T (gnd) R (-48 Vdc) T (gnd) 3 6 9 # 2 5 8 0 1 4 7 * 3 6 9 # 2 5 8 0 1 4 7 * FIGURE 5 Telephone call procedures When the calling party’s telephone set goes off hook (i.e., lifting the handset off the cradle), the switch hook in the telephone set is released, completing a dc path between the tip and the ring of the loop through the microphone. The ESS machine senses a dc current in the loop and recognizes this as an off-hook condition. This procedure is referred to as loop start operation since the loop is completed through the telephone set. The amount of dc current produced depends on the wire resistance, which varies with loop length, wire gauge, type of wire, and the impedance of the subscriber’s telephone. Typical loop resis- tance ranges from a few ohms up to approximately 1300 ohms, and typical telephone set impedances range from 500 ohms to 1000 ohms. Completing a local telephone call between two subscribers connected to the same telephone switch is accomplished through a standard set of procedures that includes the 10 Telephone Instruments and Signals 390
- 396. steps listed next. Accessing the telephone system in this manner is known as POTS (plain old telephone service): Step 1 Calling station goes off hook. Step 2 After detecting a dc current flow on the loop, the switching machine returns an audible dial tone to the calling station, acknowledging that the caller has access to the switching machine. Step 3 The caller dials the destination telephone number using one of two methods: mechanical dial pulsing or, more likely, electronic dual-tone multifrequency (Touch-Tone) signals. Step 4 When the switching machine detects the first dialed number, it removes the dial tone from the loop. Step 5 The switch interprets the telephone number and then locates the local loop for the destination telephone number. Step 6 Before ringing the destination telephone, the switching machine tests the destination loop for dc current to see if it is idle (on hook) or in use (off hook).At the same time, the switching machine locates a signal path through the switch between the two local loops. Step 7a If the destination telephone is off hook, the switching machine sends a sta- tion busy signal back to the calling station. Step 7b If the destination telephone is on hook, the switching machine sends a ring- ing signal to the destination telephone on the local loop and at the same time sends a ring back signal to the calling station to give the caller some assur- ance that something is happening. Step 8 When the destination answers the telephone, it completes the loop, causing dc current to flow. Step 9 The switch recognizes the dc current as the station answering the telephone. At this time, the switch removes the ringing and ring-back signals and com- pletes the path through the switch, allowing the calling and called parties to begin their conversation. Step 10 When either end goes on hook, the switching machine detects an open cir- cuit on that loop and then drops the connections through the switch. Placing telephone calls between parties connected to different switching machines or between parties separated by long distances is somewhat more complicated. 5 CALL PROGRESS TONES AND SIGNALS Call progress tones and call progress signals are acknowledgment and status signals that ensure the processes necessary to set up and terminate a telephone call are completed in an orderly and timely manner. Call progress tones and signals can be sent from machines to machines, machines to people, and people to machines. The people are the sub- scribers (i.e., the calling and the called party), and the machines are the electronic switching systems in the telephone offices and the telephone sets themselves. When a switching machine outputs a call progress tone to a subscriber, it must be audible and clearly identifiable. Signaling can be broadly divided into two major categories: station signaling and interoffice signaling. Station signaling is the exchange of signaling messages over local loops between stations (telephones) and telephone company switching machines. On the other hand, interoffice signaling is the exchange of signaling messages between switching ma- chines. Signaling messages can be subdivided further into one of four categories: alerting, Telephone Instruments and Signals 391
- 397. Table 1 Call Progress Tone Summary Tone or Signal Frequency Duration/Range Dial tone 350 Hz plus 440 Hz Continuous DTMF 697 Hz, 770 Hz, 852 Hz, 941 Hz, Two of eight tones 1209 Hz, 1336 Hz, 1477 Hz, 1633 Hz On, 50-ms minimum Off, 45-ms minimum, 3-s maximum MF 700 Hz, 900 Hz, 1100 Hz, Two of six tones 1300 Hz, 1500 Hz, 1700 Hz On, 90-ms minimum, 120-ms maximum Dial pulses Open/closed switch On, 39 ms Off, 61 ms Station busy 480 Hz plus 620 Hz On, 0.5 s Off, 0.5 s Equipment busy 480 Hz plus 620 Hz On, 0.2 s Off, 0.3 s Ringing 20 Hz, 90 vrms (nominal) On, 2 s Off, 4 s Ring-back 440 Hz plus 480 Hz On, 2 s Off, 4 s Receiver on hook Open loop Indefinite Receiver off hook dc current 20-mA minimum, 80-mA maximum, Receiver-left-off- 1440 Hz, 2060 Hz, 2450 Hz, 2600 Hz On, 0.1 s hook alert Off, 0.1 s Table 2 Call Progress Tone Direction of Propagation Tone or Signal Direction Dial tone Telephone office to calling station DTMF Calling station to telephone office MF Telephone office to telephone office Dial pulses Calling station to telephone office Station busy Telephone office to calling subscriber Equipment busy Telephone office to calling subscriber Ringing Telephone office to called subscriber Ring-back Telephone office to calling subscriber Receiver on hook Calling subscriber to telephone office Receiver off hook Calling subscriber to telephone office Receiver-left-off-hook alert Telephone office to calling subscriber supervising, controlling, and addressing. Alerting signals indicate a request for service, such as going off hook or ringing the destination telephone. Supervising signals provide call status information, such as busy or ring-back signals. Controlling signals provide informa- tion in the form of announcements, such as number changed to another number, a number no longer in service, and so on. Addressing signals provide the routing information, such as calling and called numbers. Examples of essential call progress signals are dial tone, dual tone multifrequency tones, multifrequency tones, dial pulses, station busy, equipment busy, ringing, ring-back, receiver on hook, and receiver off hook. Tables 1 and 2 summarize the most important call progress tones and their direction of propagation, respectively. 5-1 Dial Tone Siemens Company first introduced dial tone to the public switched telephone network in Germany in 1908. However, it took several decades before being accepted in the United Telephone Instruments and Signals 392
- 398. 1 ABC 2 DEF 3 A GHI 4 JKL 5 MNO 6 B 1209 Hz 1336 Hz Column (high-group frequencies) Row (low-group frequencies) 1477 Hz 1633 Hz PRS 7 TUV 8 WXY 9 C * 0 # D 941 Hz (Optional) 852 Hz 770 Hz 697 Hz FIGURE 6 DTMF keypad layout and frequency allocation States. Dial tone is an audible signal comprised of two frequencies: 350 Hz and 440 Hz. The two tones are linearly combined and transmitted simultaneously from the central office switching machine to the subscriber in response to the subscriber going off hook. In essence, dial tone informs subscribers that they have acquired access to the electronic switching machine and can now dial or use Touch-Tone in a destination telephone number. After a subscriber hears the dial tone and begins dialing, the dial tone is removed from the line (this is called breaking dial tone). On rare occasions, a subscriber may go off hook and not receive dial tone. This condition is appropriately called no dial tone and occurs when there are more subscribers requesting access to the switching machine than the switching machine can handle at one time. 5-2 Dual-Tone MultiFrequency Dual-tone multifrequency (DTMF) was first introduced in 1963 with 10 buttons in Western Electric 1500-type telephones. DTMF was originally called Touch-Tone. DTMF is a more efficient means than dial pulsing for transferring telephone numbers from a subscriber’s lo- cation to the central office switching machine. DTMF is a simple two-of-eight encoding scheme where each digit is represented by the linear addition of two frequencies. DTMF is strictly for signaling between a subscriber’s location and the nearest telephone office or message switching center. DTMF is sometimes confused with another two-tone signaling system called multifrequency signaling (MF), which is a two-of-six code designed to be used only to convey information between two electronic switching machines. Figure 6 shows the four-row-by-four-column keypad matrix used with a DTMF key- pad. As the figure shows, the keypad is comprised of 16 keys and eight frequencies. Most household telephones, however, are not equipped with the special-purpose keys located in the fourth column (i.e., the A, B, C, and D keys). Therefore, most household telephones ac- tually use two-of-seven tone encoding scheme. The four vertical frequencies (called the low group frequencies) are 697 Hz, 770 Hz, 852 Hz, and 941 Hz, and the four horizon- tal frequencies (called the high group frequencies) are 1209 Hz, 1336 Hz, 1477 Hz, and 1633 Hz. The frequency tolerance of the oscillators is .5%. As shown in Figure 6, the digits 2 through 9 can also be used to represent 24 of the 26 letters (Q and Z are omitted). The letters were originally used to identify one local telephone exchange from another, Telephone Instruments and Signals 393
- 399. Table 3 DTMF Specifications Transmitter Receiver (Subscriber) Parameter (Local Office) 10 dBm Minimum power level (single frequency) 25 dBm 2 dBm Maximum power level (two tones) 0 dBm 4 dB Maximum power difference between two tones 4 dB 50 ms Minimum digit duration 40 ms 45 ms Minimum interdigit duration 40 ms 3 s Maximum interdigit time period 3 s Maximum echo level relative to transmit frequency level (10 dB) Maximum echo delay (20 ms) such as BR for Bronx, MA for Manhattan, and so on. Today, the letters are used to person- alize telephone numbers. For example; 1-800-UPS-MAIL equates to the telephone number 1-800-877-6245. When a digit (or letter) is selected, two of the eight frequencies (or seven for most home telephones) are transmitted (one from the low group and one from the high group). For example, when the digit 5 is depressed, 770 Hz and 1336 Hz are transmitted si- multaneously. The eight frequencies were purposely chosen so that there is absolutely no harmonic relationship between any of them, thus eliminating the possibility of one fre- quency producing a harmonic that might be misinterpreted as another frequency. The major advantages for the subscriber in using Touch-Tone signaling over dial pulsing is speed and control. With Touch-Tone signaling, all digits (and thus telephone numbers) take the same length of time to produce and transmit. Touch-Tone signaling also eliminates the impulse noise produced from the mechanical switches necessary to produce dial pulses. Probably the most important advantage of DTMF over dial pulsing is the way in which the telephone company processes them. Dial pulses cannot pass through a central office exchange (local switching machine), whereas DTMF tones will pass through an ex- change to the switching system attached to the called number. Table 3 lists the specifications for DTME. The transmit specifications are at the sub- scriber’s location, and the receive specifications are at the local switch. Minimum power lev- els are given for a single frequency, and maximum power levels are given for two tones. The minimum duration is the minimum time two tones from a given digit must remain on. The in- terdigit time specifies the minimum and maximum time between the transmissions of any two successivedigits.Anechooccurswhenapairoftonesisnottotallyabsorbedbythelocalswitch and a portion of the power is returned to the subscriber. The maximum power level of an echo is 10 dB below the level transmitted by the subscriber and must be delayed less than 20 ms. 5-3 Multifrequency Multifrequency tones (codes) are similar to DTMF signals in that they involve the simulta- neous transmission of two tones. MF tones are used to transfer digits and control signals between switching machines, whereas DTMF signals are used to transfer digits and control signals between telephone sets and local switching machines. MF tones are combinations of two frequencies that fall within the normal speech bandwidth so that they can be propa- gated over the same circuits as voice. This is called in-band signaling. In-band signaling is rapidly being replaced by out-of-band signaling. MF codes are used to send information between the control equipment that sets up connections through a switch when more than one switch is involved in completing a call. MF codes are also used to transmit the calling and called numbers from the originating tele- phone office to the destination telephone office. The calling number is sent first, followed by the called number. Table 4 lists the two-tone MF combinations and the digits or control information they represent. As the table shows, MF tones involve the transmission of two-of-six possi- Telephone Instruments and Signals 394
- 400. Table 4 Multifrequency Codes Frequencies (Hz) Digit or Control 700 900 1 700 1100 2 700 1300 4 700 1500 7 900 1100 3 900 1300 5 900 1500 8 1100 1300 6 1100 1500 9 1100 1700 Key pulse (KP) 1300 1500 0 1500 1700 Start (ST) 2600 Hz IDLE Off hook closed loop current flowing On hook open loop no current Switching machine returns dial tone 39 ms Make 61 ms Break Next digit sequence of pulses Dial pulse period 100 ms 100 ms Digit 3 Interdigit period 300 ms minimum 100 ms FIGURE 7 Dial pulsing sequence ble frequencies representing the 10 digits plus two control signals. The six frequencies are 700 Hz, 900 Hz, 1100 Hz, 1300 Hz, 1500 Hz, and 1700 Hz. Digits are transmitted at a rate of seven per second, and each digit is transmitted as a 68-ms burst. The key pulse (KP) sig- nal is a multifrequency control tone comprised of 1100 Hz plus 1700 Hz, ranging from 90 ms to 120 ms. The KP signal is used to indicate the beginning of a sequence of MF digits. The start (ST) signal is a multifrequency control tone used to indicate the end of a sequence of dialed digits. From the perspective of the telephone circuit, the ST control signal indi- cates the beginning of the processing of the signal. The IDLE signal is a 2600-Hz single- frequency tone placed on a circuit to indicate the circuit is not currently in use. For exam- ple, KP 3 1 5 7 3 6 1 0 5 3 ST is the sequence transmitted for the telephone number 315-736-1053. 5-4 Dial Pulses Dial pulsing (sometimes called rotary dial pulsing) is the method originally used to trans- fer digits from a telephone set to the local switch. Pulsing digits from a rotary switch began soon after the invention of the automatic switching machine. The concept of dial pulsing is quite simple and is depicted in Figure 7. The process begins when the telephone set is lifted off hook, completing a path for current through the local loop. When the switching machine detects the off-hook condition, it responds with dial tone. After hearing the dial tone, the subscriber begins dial pulsing digits by rotating a mechanical dialing mechanism Telephone Instruments and Signals 395
- 401. and then letting it return to its rest position. As the rotary switch returns to its rest position, it outputs a series of dial pulses corresponding to the digit dialed. When a digit is dialed, the loop circuit alternately opens (breaks) and closes (makes) a prescribed number of times. The number of switch make/break sequences corresponds to the digit dialed (i.e., the digit 3 produces three switch openings and three switch closures). Dial pulses occur at 10 make/break cycles per second (i.e., a period of 100 ms per pulse cy- cle). For example, the digit 5 corresponds to five make/break cycles lasting a total of 500 ms. The switching machine senses and counts the number of make/break pairs in the se- quence. The break time is nominally 61 ms, and the make time is nominally 39 ms. Digits are separated by an idle period of 300 ms called the interdigit time. It is essential that the switching machine recognize the interdigit time so that it can separate the pulses from suc- cessive digits. The central office switch incorporates a special time-out circuit to ensure that the break part of the dialing pulse is not misinterpreted as the phone being returned to its on-hook (idle) condition. All digits do not take the same length of time to dial. For example, the digit 1 requires only one make/break cycle, whereas the digit 0 requires 10 cycles. Therefore, all telephone numbers do not require the same amount of time to dial or to transmit. The minimum time to dial pulse out the seven-digit telephone number 987-1234 is as follows: where ID is the interdigit time (300 ms) and the total minimum time is 5200 ms, or 5.2 seconds. 5-5 Station Busy In telephone terminology, a station is a telephone set. A station busy signal is sent from the switching machine back to the calling station whenever the called telephone number is off hook (i.e., the station is in use). The station busy signal is a two-tone signal comprised of 480 Hz and 620 Hz. The two tones are on for 0.5 seconds, then off for 0.5 seconds. Thus, a busy signal repeats at a 60-pulse-per-minute (ppm) rate. 5-6 Equipment Busy The equipment busy signal is sometimes called a congestion tone or a no-circuits-available tone.The equipment busy signal is sent from the switching machine back to the calling station wheneverthesystemcannotcompletethecallbecauseofequipmentunavailability(i.e.,allthe circuits, switches, or switching paths are already in use). This condition is called blocking and occurs whenever the system is overloaded and more calls are being placed than can be com- pleted. The equipment busy signal uses the same two frequencies as the station busy signal, except the equipment busy signal is on for 0.2 seconds and off for 0.3 seconds (120 ppm). Be- cause an equipment busy signal repeats at twice the rate as a station busy signal, an equipment busy is sometimes called a fast busy, and a station busy is sometimes called a slow busy. The telephone company refers to an equipment busy condition as a can’t complete. 5-7 Ringing The ringing signal is sent from a central office to a subscriber whenever there is an incom- ing call. The purpose of the ringing signal is to ring the bell in the telephone set to alert the subscriber that there is an incoming call. If there is no bell in the telephone set, the ringing signal is used to trigger another audible mechanism, which is usually a tone oscillator cir- cuit. The ringing signal is nominally a 20-Hz, 90-Vrms signal that is on for 2 seconds and then off for 4 seconds. The ringing signal should not be confused with the actual ringing sound the bell makes. The audible ring produced by the bell was originally made as an- noying as possible so that the called end would answer the telephone as soon as possible, thus tying up common usage telephone equipment in the central office for the minimum length of time. digit 9 ID 8 ID 7 ID 1 ID 2 ID 3 ID 4 time 1ms2 900 300 800 300 700 300 100 300 200 300 300 300 400 Telephone Instruments and Signals 396
- 402. 5-8 Ring-Back The ring-back signal is sent back to the calling party at the same time the ringing signal is sent to the called party. However, the ring and ring-back signals are two distinctively dif- ferent signals. The purpose of the ring-back signal is to give some assurance to the calling party that the destination telephone number has been accepted, processed, and is being rung. The ring-back signal is an audible combination of two tones at 440 Hz and 480 Hz that are on for 2 seconds and then off for 4 seconds. 5-9 Receiver On/Off Hook When a telephone is on hook, it is not being used, and the circuit is in the idle (or open) state. The term on hook was derived in the early days of telephone when the telephone hand- set was literally placed on a hook (the hook eventually evolved into a cradle). When the tele- phone set is on hook, the local loop is open, and there is no current flowing on the loop. An on-hook signal is also used to terminate a call and initiate a disconnect. When the telephone set is taken off hook, a switch closes in the telephone that com- pletes a dc path between the two wires of the local loop. The switch closure causes a dc cur- rent to flow on the loop (nominally between 20 mA and 80 mA, depending on loop length and wire gauge). The switching machine in the central office detects the dc current and rec- ognizes it as a receiver off-hook condition (sometimes called a seizure or request for service). The receiver off-hook condition is the first step to completing a telephone call. The switch- ing machine will respond to the off-hook condition by placing an audible dial tone on the loop. The off-hook signal is also used at the destination end as an answer signal to indicate that the called party has answered the telephone. This is sometimes referred to as a ring trip because when the switching machine detects the off-hook condition, it removes (or trips) the ringing signal. 5-10 Other Nonessential Signaling and Call Progress Tones There are numerous additional signals relating to initiating, establishing, completing, and terminating a telephone call that are nonessential, such as call waiting tones, caller waiting tones, calling card service tones, comfort tones, hold tones, intrusion tones, stutter dial tone (for voice mail), and receiver off-hook tones (also called howler tones). 6 CORDLESS TELEPHONES Cordless telephones are simply telephones that operate without cords attached to the hand- set. Cordless telephones originated around 1980 and were quite primitive by today’s stan- dards. They originally occupied a narrow band of frequencies near 1.7 MHz, just above the AM broadcast band, and used the 117-vac, 60-Hz household power line for an antenna. These early units used frequency modulation (FM) and were poor quality and susceptible to interference from fluorescent lights and automobile ignition systems. In 1984, the FCC reallocated cordless telephone service to the 46-MHz to 49-MHz band. In 1990, the FCC extended cordless telephone service to the 902-MHz to 928-MHz band, which appreciated a superior signal-to-noise ratio. Cordless telephone sets transmit and receive over narrow- band FM (NBFM) channels spaced 30 kHz to 100 kHz apart, depending on the modulation and frequency band used. In 1998, the FCC expanded service again to the 2.4-GHz band. Adaptive differential pulse code modulation and spread spectrum technology (SST) are used exclusively in the 2.4-GHz band, while FM and SST digital modulation are used in the 902-MHz to 928-MHz band. Digitally modulated SST telephones offer higher quality and more security than FM telephones. In essence, a cordless telephone is a full-duplex, battery-operated, portable radio transceiver that communicates directly with a stationary transceiver located somewhere in Telephone Instruments and Signals 397
- 403. Antenna Speaker Microphone Portable Cordless Telephone Set Two-way Radio Wave Propagation Wall Jack Base Station Unit Block Diagram Block Diagram AC Power To local loop Transmitter Receiver + – Keyboard Antenna Transmitter Telco Interface Receiver Battery- powered Power Supply FIGURE 8 Cordless telephone system the subscriber’s home or office. The basic layout for a cordless telephone is shown in Figure 8. The base station is an ac-powered stationary radio transceiver (transmitter and receiver) connected to the local loop through a cord and telephone company interface unit. The inter- face unit functions in much the same way as a standard telephone set in that its primary func- tion is to interface the cordless telephone with the local loop while being transparent to the user. Therefore, the base station is capable of transmitting and receiving both supervisory and voice signals over the subscriber loop in the same manner as a standard telephone. The base station must also be capable of relaying voice and control signals to and from the portable telephone set through the wireless transceiver. In essence, the portable telephone set is a battery-powered, two-way radio capable of operating in the full-duplex mode. Because a portable telephone must be capable of communicating with the base station in the full-duplex mode, it must transmit and receive at different frequencies. In 1984, the FCC al- located 10 full-duplex channels for 46-MHz to 49-MHz units. In 1995 to help relieve conges- tion, the FCC added 15 additional full-duplex channels and extended the frequency band to in- clude frequencies in the 43-MHz to 44-MHz band. Base stations transmit on high-band frequenciesandreceiveonlow-bandfrequencies,whiletheportableunittransmitsonlow-band frequencies and receives on high-band frequencies. The frequency assignments are listed in Table 5. Channels 16 through 25 are the original 10 full-duplex carrier frequencies. The maxi- mum transmit power for both the portable unit and the base station is 500 mW. This stipulation limits the useful range of a cordless telephone to within 100 feet or less of the base station. Telephone Instruments and Signals 398
- 404. Telephone Instruments and Signals Table 5 43-MHz- to 49-MHz-Band Cordless Telephone Frequencies Portable Unit Channel Transmit Frequency (MHz) Receive Frequency (MHz) 1 43.720 48.760 2 43.740 48.840 3 43.820 48.860 4 43.840 48.920 5 43.920 49.920 6 43.960 49.080 7 44.120 49.100 8 44.160 49.160 9 44.180 49.200 10 44.200 49.240 11 44.320 49.280 12 44.360 49.360 13 44.400 49.400 14 44.460 49.460 15 44.480 49.500 16 46.610 49.670 17 46.630 49.845 18 46.670 49.860 19 46.710 49.770 20 46.730 49.875 21 46.770 49.830 22 46.830 49.890 23 46.870 49.930 24 46.930 49.970 25 46.970 49.990 Note. Base stations transmit on the 49-MHz band and receive on the 46-MHz band. Cordless telephones using the 2.4-GHz band offer excellent sound quality utilizing digital modulation and twin-band transmission to extend their range. With twin-band trans- mission, base stations transmit in the 2.4-GHz band, while portable units transmit in the 902-MHz to 928-MHz band. 7 CALLER ID Caller ID (identification) is a service originally envisioned by ATT in the early 1970s, al- though local telephone companies have only recently offered it. The basic concept of caller ID is quite simple. Caller ID enables the destination station of a telephone call to display the name and telephone number of the calling party before the telephone is answered (i.e., while the telephone is ringing). This allows subscribers to screen incoming calls and decide whether they want to answer the telephone. The caller ID message is a simplex transmission sent from the central office switch over the local loop to a caller ID display unit at the destination station (no response is provided). The caller ID information is transmitted and received using Bell System 202-compatible modems (ITU V.23 standard). This standard specifies a 1200-bps FSK (frequency shift key- ing) signal with a 1200-Hz mark frequency (fm) and a 2200-Hz space frequency (fm). The FSK signal is transmitted in a burst between the first and second 20-Hz, 90-Vrms ringing signals, as shown in Figure 9a. Therefore, to ensure detection of the caller ID signal, the telephone must ring at least twice before being answering. The caller ID signal does not begin until 500 ms after the end of the first ring and must end 500 ms before the beginning of the second ring. Therefore, the caller ID signal has a 3-second window in which it must be transmitted. 399
- 405. Telephone Instruments and Signals 1st 20 Hz, 90 vrms ring 2nd 20 Hz, 90 vrms ring Caller ID signal 1200 bps FSK fm = 1200 Hz fs = 2200 Hz (a) 2 seconds 2 seconds 4 seconds 3 second window 0.5 seconds 0.5 seconds Channel Seizure 200 ms Alternating 1/0 Sequence (55H or AAH) Month 10 bits Month 10 bits Day 10 bits Day 10 bits Hour 10 bits Hour 10 bits Minute 10 bits Minute 10 bits Caller's Name and Telephone Number Conditioning Signal 130 ms all 1s Message Type 8.33 ms Message Length 8.33 ms Check Sum 6.67 ms Caller ID Data Field 240 bits 156 bits 10 bits 10 bits (8 data bits, one start bit, and one stop bit) 8 bits Variable length (b) Caller ID Signal 66.7 ms Variable Length (8 ASCII data bits, one start bit, and one stop bit) (ASCII characters) FIGURE 9 Caller ID: (a) ringing cycle; (b) frame format The format for a caller ID signal is shown in Figure 9b. The 500-ms delay after the first ringing signal is immediately followed by the channel seizure field, which is a 200- ms-long sequence of alternating logic 1s and logic 0s (240 bits comprised of 120 pairs of alternating 1/0 bits, either 55 hex or AA hex). A conditioning signal field immediately fol- lows the channel seizure field. The conditioning signal is a continuous 1200-Hz tone last- ing for 130 ms, which equates to 156 consecutive logic 1 bits. The protocol used for the next three fields—message type field, message length field, and caller ID data field—specifies asynchronous transmission of 16-bit characters (with- out parity) framed by one start bit (logic 0) and one stop bit (logic 1) for a total of 10 bits per character. The message type field is comprised of a 16-bit hex code, indicating the type of service and capability of the data message. There is only one message type field currently used with caller ID (04 hex). The message type field is followed by a 16-bit message length field, which specifies the total number of characters (in binary) included in the caller ID data field. For example, a message length code of 15 hex (0001 0101) equates to the num- ber 21 in decimal. Therefore, a message length code of 15 hex specifies 21 characters in the caller ID data field. The caller ID data field uses extended ASCII coded characters to represent a month code (01 through 12), a two-character day code (01 through 31), a two-character hour code in local military time (00 through 23), a two-character minute code (00 through 59), and a variable-length code, representing the caller’s name and telephone number. ASCII coded digits are comprised of two independent hex characters (eight bits each). The first hex char- acter is always 3 (0011 binary), and the second hex character represents a digit between 0 and 9 (0000 to 1001 binary). For example, 30 hex (0011 0000 binary) equates to the digit 0, 31 hex (0011 0001 binary) equates to the digit 1, 39 hex (0011 1001) equates to the digit 400
- 406. Telephone Instruments and Signals 9, and so on. The caller ID data field is followed by a checksum for error detection, which is the 2’s complement of the module 256 sum of the other words in the data message (mes- sage type, message length, and data words). Example 1 Interpret the following hex code for a caller ID message (start and stop bits are not included in the hex codes): 04 12 31 31 32 37 31 35 35 37 33 31 35 37 33 36 31 30 35 33 xx Solution 04—message type word 12—18 decimal (18 characters in the caller ID data field) 31, 31—ASCII code for 11 (the month of November) 32, 37—ASCII code for 27 (the 27th day of the month) 31, 35—ASCII code for 15 (the 15th hour—3:00 P.M.) 35, 37—ASCII code for 57 (57 minutes after the hour—3:57 P.M.) 33, 31, 35, 37, 33, 36, 31, 30, 35, 33—10-digit ASCII-coded telephone number (315 736–1053) xx—checksum (00 hex to FF hex) 8 ELECTRONIC TELEPHONES Although 500- and 2500-type telephone sets still work with the public telephone network, they are becoming increasingly more difficult to find. Most modern-day telephones have replaced many of the mechanical functions performed in the old telephone sets with elec- tronic circuits. Electronic telephones use integrated-circuit technology to perform many of the basic telephone functions as well as a myriad of new and, and in many cases, nonessen- tial functions. The refinement of microprocessors has also led to the development of multiple-line, full-feature telephones that permit automatic control of the telephone set’s features, including telephone number storage, automatic dialing, redialing, and caller ID. However, no matter how many new gadgets are included in the new telephone sets, they still have to interface with the telephone network in much the same manner as telephones did a century ago. Figure 10 shows the block diagram for a typical electronic telephone comprised of one multifunctional integrated-circuit chip, a microprocessor chip, a Touch-Tone keypad, a speaker, a microphone, and a handful of discrete devices. The major components included in the multifunctional integrated circuit chip are DTMF tone generator, MPU (micro- processor unit) interface circuitry, random access memory (RAM), tone ringer circuit, speech network, and a line voltage regulator. The Touch-Tone keyboard provides a means for the operator of the telephone to ac- cess the DTMF tone generator inside the multifunction integrated-circuit chip. The exter- nal crystal provides a stable and accurate frequency reference for producing the dual-tone multifrequency signaling tones. The tone ringer circuit is activated by the reception of a 20-Hz ringing signal. Once the ringing signal is detected, the tone ringer drives a piezoelectric sound element that pro- duces an electronic ring (without a bell). The voltage regulator converts the dc voltage received from the local loop and con- verts it to a constant-level dc supply voltage to operate the electronic components in the telephone. The internal speech network contains several amplifiers and associated compo- nents that perform the same functions as the hybrid did in a standard telephone. The microprocessor interface circuit interfaces the MPU to the multifunction chip. The MPU, with its internal RAM, controls many of the functions of the telephone, such as 401
- 407. Telephone Instruments and Signals number storage, speed dialing, redialing, and autodialing. The bridge rectifier protects the telephone from the relatively high-voltage ac ringing signal, and the switch hook is a me- chanical switch that performs the same functions as the switch hook on a standard tele- phone set. 9 PAGING SYSTEMS Most paging systems are simplex wireless communications system designed to alert sub- scribers of awaiting messages. Paging transmitters relay radio signals and messages from wire-line and cellular telephones to subscribers carrying portable receivers. The simplified block diagram of a paging system is shown in Figure 11. The infrastructure used with pag- ing systems is somewhat different than the one used for cellular telephone system. This is because standard paging systems are one way, with signals transmitted from the paging sys- tem to portable pager and never in the reverse direction. There are narrow-, mid-, and wide- area pagers (sometimes called local, regional, and national). Narrow-area paging systems operate only within a building or building complex, mid-area pagers cover an area of sev- eral square miles, and wide-area pagers operate worldwide. Most pagers are mid-area where one centrally located high-power transmitter can cover a relatively large geographic area, typically between 6 and 10 miles in diameter. To contact a person carrying a pager, simply dial the telephone number assigned that person’s portable pager. The paging company receives the call and responds with a query requesting the telephone number you wish the paged person to call. After the number is en- tered, a terminating signal is appended to the number, which is usually the # sign. The caller then hangs up. The paging system converts the telephone number to a digital code and trans- mits it in the form of a digitally encoded signal over a wireless communications system. The signal may be simultaneously sent from more than one radio transmitter (sometimes called simulcasting or broadcasting), as is necessary in a wide-area paging system. If the paged person is within range of a broadcast transmitter, the targeted pager will receive the message. The message includes a notification signal, which either produces an audible beep or causes the pager to vibrate and the number the paged unit should call shown on an al- phanumeric display. Some newer paging units are also capable of displaying messages as well as the telephone number of the paging party. RJ11 To Local Loop Touch tone keyboard MPU Piezoelectric sound element Loop interface 1 2 3 A 4 5 6 B 7 8 9 C * 0 # D Tone ringer circuit Line voltage regulator circuit Crystal reference DTMF Circuit Speech network MPU interface circuit Multifunction IC Chip Speaker Microphone FIGURE 10 Electronic telephone set 402
- 408. Telephone Instruments and Signals Local telephone office Wireline telephone Cellular telephone Portable cordless telephone Cellular telephone office Cordless telephone base station Paging service office Radio transmitter Portable pager 3 6 9 # 2 5 8 0 1 4 7 * FIGURE 11 Simplified block diagram of a standard simplex paging system Early paging systems used FM; however, most modern paging systems use FSK or PSK. Pagers typically transmit bit rates between 200 bps and 6400 bps with the following carrier frequency bands: 138 MHz to 175 MHz, 267 MHz to 284 MHz, 310 MHz to 330 MHz, 420 MHz to 470 MHz, and several frequency slots within the 900-MHz band. Each portable pager is assigned a special code, called a cap code, which includes a sequence of digits or a combination of digits and letters. The cap code is broadcasted along with the paging party’s telephone number. If the portable paging unit is within range of the broadcasting transmitter, it will receive the signal, demodulate it, and recognize its cap code. Once the portable pager recognizes its cap code, the callback number and perhaps a message will be displayed on the unit. Alphanumeric messages are generally limited to be- tween 20 and 40 characters in length. Early paging systems, such as one developed by the British Post Office called Post Office Code Standardization Advisory Group (POCSAG), transmitted a two-level FSK sig- nal. POCSAG used an asynchronous protocol, which required a long preamble for syn- chronization. The preamble begins with a long dotting sequence (sometimes called a dotting comma) to establish clock synchronization. Data rates for POCSAG are 512 bps, 1200 bps, and 2400 bps. With POCSAG, portable pagers must operate in the always-on mode all the time, which means the pager wastes much of its power resources on nondata preamble bits. In the early 1980s, the European Telecommunications Standards Institute (ETSI) de- veloped the ERMES protocol. ERMES transmitted data at a 6250 bps rate using four-level FSK (3125 baud). ERMES is a synchronous protocol, which requires less time to synchro- nize. ERMES supports 16 25-kHz paging channels in each of its frequency bands. 403
- 409. Telephone Instruments and Signals The most recent paging protocol, FLEX, was developed in the 1990s. FLEX is de- signed to minimize power consumption in the portable pager by using a synchronous time- slotted protocol to transmit messages in precise time slots. With FLEX, each frame is com- prised of 128 data frames, which are transmitted only once during a 4-minute period. Each frame lasts for 1.875 seconds and includes two synchronizing sequences, a header contain- ing frame information and pager identification addresses, and 11 discrete data blocks. Each portable pager is assigned a specific frame (called a home frame) within the frame cycle that it checks for transmitted messages. Thus, a pager operates in the high-power standby con- dition for only a few seconds every 4 minutes (this is called the wakeup time). The rest of the time, the pager is in an ultra-low power standby condition. When a pager is in the wakeup mode, it synchronizes to the frame header and then adjusts itself to the bit rate of the received signal. When the pager determines that there is no message waiting, it puts it- self back to sleep, leaving only the timer circuit active. QUESTIONS 1. Define the terms communications and telecommunications. 2. Define plain old telephone service. 3. Describe a local subscriber loop. 4. Where in a telephone system is the local loop? 5. Briefly describe the basic functions of a standard telephone set. 6. What is the purpose of the RJ-11 connector? 7. What is meant by the terms tip and ring? 8. List and briefly describe the essential components of a standard telephone set. 9. Briefly describe the steps involved in completing a local telephone call. 10. Explain the basic purpose of call progress tones and signals. 11. List and describe the two primary categories of signaling. 12. Describe the following signaling messages: alerting, supervising, controlling, and addressing. 13. What is the purpose of dial tone, and when is it applied to a telephone circuit? 14. Briefly describe dual-tone multifrequency and multifrequency signaling and tell where they are used. 15. Describe dial pulsing. 16. What is the difference between a station busy signal and an equipment busy signal? 17. What is the difference between a ringing and a ring-back signal? 18. Briefly describe what happens when a telephone set is taken off hook. 19. Describe the differences between the operation of a cordless telephone and a standard telephone. 20. Explain how caller ID operates and when it is used. 21. Briefly describe how a paging system operates. 404
- 410. The Telephone Circuit CHAPTER OUTLINE 1 Introduction 4 Units of Power Measurement 2 The Local Subscriber Loop 5 Transmission Parameters and Private-Line Circuits 3 Telephone Message–Channel Noise 6 Voice-Frequency Circuit Arrangements and Noise Weighting 7 Crosstalk OBJECTIVES ■ Define telephone circuit, message, and message channel ■ Describe the transmission characteristics a local subscriber loop ■ Describe loading coils and bridge taps ■ Describe loop resistance and how it is calculated ■ Explain telephone message–channel noise and C-message noise weighting ■ Describe the following units of power measurement: db, dBm, dBmO, rn, dBrn, dBrnc, dBrn 3-kHz flat, and dBrncO ■ Define psophometric noise weighting ■ Define and describe transmission parameters ■ Define private-line circuit ■ Explain bandwidth, interface, and facilities parameters ■ Define line conditioning and describe C- and D-type conditioning ■ Describe two-wire and four-wire circuit arrangements ■ Explain hybrids, echo suppressors, and echo cancelers ■ Define crosstalk ■ Describe nonlinear, transmittance, and coupling crosstalk From Chapter 9 of Advanced Electronic Communications Systems, Sixth Edition. Wayne Tomasi. Copyright © 2004 by Pearson Education, Inc. Published by Pearson Prentice Hall. All rights reserved. 405
- 411. 1 INTRODUCTION A telephone circuit is comprised of two or more facilities, interconnected in tandem, to pro- vide a transmission path between a source and a destination. The interconnected facilities may be temporary, as in a standard telephone call, or permanent, as in a dedicated private- line telephone circuit. The facilities may be metallic cable pairs, optical fibers, or wireless carrier systems. The information transferred is called the message, and the circuit used is called the message channel. Telephone companies offer a wide assortment of message channels ranging from a basic 4-kHz voice-band circuit to wideband microwave, satellite, or optical fiber transmis- sion systems capable of transferring high-resolution video or wideband data. The follow- ing discussion is limited to basic voice-band circuits. In telephone terminology, the word message originally denoted speech information. However, this definition has been ex tended to include any signal that occupies the same bandwidth as a standard voice channel. Thus, a message channel may include the transmission of ordinary speech, supervisory sig- nals, or data in the form of digitally modulated carriers (FSK, PSK, QAM, and so on). The network bandwidth for a standard voice-band message channel is 4 kHz; however, a por- tion of that bandwidth is used for guard bands and signaling. Guard bands are unused fre- quency bands located between information signals. Consequently, the effective channel bandwidth for a voice-band message signal (whether it be voice or data) is approximately 300 Hz to 3000 Hz. 2 THE LOCAL SUBSCRIBER LOOP The local subscriber loop is the only facility required by all voice-band circuits, as it is the means by which subscriber locations are connected to the local telephone company. In essence, the sole purpose of a local loop is to provide subscribers access to the public tele- phone network. The local loop is a metallic transmission line comprised of two insulated copper wires (a pair) twisted together. The local loop is the primary cause of attenuation and phase distortion on a telephone circuit. Attenuation is an actual loss of signal strength, and phase distortion occurs when two or more frequencies undergo different amounts of phase shift. The transmission characteristics of a cable pair depend on the wire diameter, con- ductor spacing, dielectric constant of the insulator separating the wires, and the conductiv- ity of the wire. These physical properties, in turn, determine the inductance, resistance, ca- pacitance, and conductance of the cable. The resistance and inductance are distributed along the length of the wire, whereas the conductance and capacitance exist between the two wires. When the insulation is sufficient, the effects of conductance are generally neg- ligible. Figure 1a shows the electrical model for a copper-wire transmission line. The electrical characteristics of a cable (such as inductance, capacitance, and resis- tance) are uniformly distributed along its length and are appropriately referred to as distributed parameters. Because it is cumbersome working with distributed parameters, it is common practice to lump them into discrete values per unit length (i.e., millihenrys per mile, microfarads per kilometer, or ohms per 1000 feet). The amount of attenuation and phase delay experienced by a signal propagating down a metallic transmission line is a function of the frequency of the signal and the electrical characteristics of the cable pair. There are seven main component parts that make up a traditional local loop: Feeder cable (F1). The largest cable used in a local loop, usually 3600 pair of copper wire placed underground or in conduit. Serving area interface (SAI). A cross-connect point used to distribute the larger feeder cable into smaller distribution cables. The Telephone Circuit 406
- 412. FIGURE 1 (a) Electrical model of a copper-wire transmission line, (b) frequency-versus-attenuation characteristics for unloaded and loaded cables Distribution cable (F2). A smaller version of a feeder cable containing less wire pairs. Subscriber or standard network interface (SNI). A device that serves as the demar- cation point between local telephone company responsibility and subscriber respon- sibility for telephone service. Drop wire. The final length of cable pair that terminates at the SNI. Aerial. That portion of the local loop that is strung between poles. Distribution cable and drop-wire cross-connect point. The location where individual cable pairs within a distribution cable are separated and extended to the subscriber’s location on a drop wire. Two components often found on local loops are loading coils and bridge taps. The Telephone Circuit 407
- 413. 2-1 Loading Coils Figure 1b shows the effect of frequency on attenuation for a 12,000-foot length of 26-gauge copper cable. As the figure shows, a 3000-Hz signal experiences 6 dB more attenuation than a 500-Hz signal on the same cable. In essence, the cable acts like a low-pass filter. Extensive studies of attenuation on cable pairs have shown that a substantial reduction in attenuation is achieved by increasing the inductance value of the cable. Minimum attenuation requires a value of inductance nearly 100 times the value obtained in ordinary twisted-wire cable. Achieving such values on a uniformly distributed basis is impractical. Instead, the desired ef- fect can be obtained by adding inductors periodically in series with the wire. This practice is called loading, and the inductors are called loading coils. Loading coils placed in a cable de- crease the attenuation, increase the line impedance, and improve transmission levels for cir- cuits longer than 18,000 feet. Loading coils allowed local loops to extend three to four times their previous length.A loading coil is simply a passive conductor wrapped around a core and placed in series with a cable creating a small electromagnet. Loading coils can be placed on telephone poles, in manholes, or on cross-connect boxes. Loading coils increase the effective distance that a signal must travel between two locations and cancels the capacitance that in- herently builds up between wires with distance. Loading coils first came into use in 1900. Loaded cables are specified by the addition of the letter codes A, B, C, D, E, F, H, X, orY, which designate the distance between loading coils and by numbers, which indicate the inductance value of the wire gauge. The letters indicate that loading coils are separated by 700, 3000, 929, 4500, 5575, 2787, 6000, 680, or 2130 feet, respectively. B-, D-, and H-type loading coils are the most common because their separations are representative of the dis- tances between manholes.The amount of series inductance added is generally 44 mH, 88 mH, or 135 mH. Thus, a cable pair designated 26H88 is made from 26-gauge wire with 88 mH of series inductance added every 6000 feet. The loss-versus-frequency characteristics for a loaded cable are relatively flat up to approximately 2000 Hz, as shown in Figure 1b. From the figure, it can be seen that a 3000-Hz signal will suffer only 1.5 dB more loss than a 500-Hz signal on 26-gauge wire when 88-mH loading coils are spaced every 6000 feet. Loading coils cause a sharp drop in frequency response at approximately 3400 Hz, which is undesirable for high-speed data transmission. Therefore, for high-performance data transmission, loading coils should be removed from the cables. The low-pass charac- teristics of a cable also affect the phase distortion-versus-frequency characteristics of a sig- nal. The amount of phase distortion is proportional to the length and gauge of the wire. Loading a cable also affects the phase characteristics of a cable. The telephone company must often add gain and delay equalizers to a circuit to achieve the minimum requirements. Equalizers introduce discontinuities or ripples in the bandpass characteristics of a circuit. Automatic equalizers in data modems are sensitive to this condition, and very often an overequalized circuit causes as many problems to a data signal as an underequalized circuit. 2-2 Bridge Taps A bridge tap is an irregularity frequently found in cables serving subscriber locations. Bridge taps are unused sections of cable that are connected in shunt to a working cable pair, such as a local loop. Bridge taps can be placed at any point along a cable’s length. Bridge taps were used for party lines to connect more than one subscriber to the same local loop. Bridge taps also increase the flexibility of a local loop by allowing the cable to go to more than one junction box, although it is unlikely that more than one of the cable pairs leaving a bridging point will be used at any given time. Bridge taps may or may not be used at some future time, depending on service demands. Bridge taps increase the flexibility of a cable by making it easier to reassign a cable to a different subscriber without requiring a person working in the field to cross connect sections of cable. Bridge taps introduce a loss called bridging loss. They also allow signals to split and propagate down more than one wire. Signals that propagate down unterminated (open- The Telephone Circuit 408
- 414. circuited) cables reflect back from the open end of the cable, often causing interference with the original signal. Bridge taps that are short and closer to the originating or terminating ends often produce the most interference. Bridge taps and loading coils are not generally harmful to voice transmissions, but if improperly used, they can literally destroy the integrity of a data signal. Therefore, bridge taps and loading coils should be removed from a cable pair that is used for data transmis- sion. This can be a problem because it is sometimes difficult to locate a bridge tap. It is es- timated that the average local loop can have as many as 16 bridge taps. 2-3 Loop Resistance The dc resistance of a local loop depends primarily on the type of wire and wire size. Most local loops use 18- to 26-gauge, twisted-pair copper wire. The lower the wire gauge, the larger the diameter, the less resistance, and the lower the attenuation. For example, 26- gauge unloaded copper wire has an attenuation of 2.67 dB per mile, whereas the same length of 19-gauge copper wire has only 1.12 dB per mile. Therefore, the maximum length of a local loop using 19-gauge wire is twice as long as a local loop using 26-gauge wire. The total attenuation of a local loop is generally limited to a maximum value of 7.5 dB with a maximum dc resistance of 1300 Ω, which includes the resistance of the tele- phone (approximately 120 Ω). The dc resistance of 26-gauge copper wire is approximately 41 Ω per 1000 feet, which limits the round-trip loop length to approximately 5.6 miles. The maximum distance for lower-gauge wire is longer of course. The dc loop resistance for copper conductors is approximated by (1) where Rdc dc loop resistance (ohms per mile) d wire diameter (inches) 3 TELEPHONE MESSAGE–CHANNEL NOISE AND NOISE WEIGHTING The noise that reaches a listener’s ears affects the degree of annoyance to the listener and, to some extent, the intelligibility of the received speech. The total noise is comprised of room background noise and noise introduced in the circuit. Room background noise on the listening subscriber’s premises reaches the ear directly through leakage around the receiver and indirectly by way of the sidetone path through the telephone set. Room noise from the talking subscriber’s premises also reaches the listener over the communications channel. Circuit noise is comprised mainly of thermal noise, nonlinear distortion, and impulse noise, which are described in a later section of this chapter. The measurement of interference (noise), like the measurement of volume, is an ef- fort to characterize a complex signal. Noise measurements on a telephone message channel are characterized by how annoying the noise is to the subscriber rather than by the absolute magnitude of the average noise power. Noise interference is comprised of two components: annoyance and the effect of noise on intelligibility, both of which are functions of fre- quency. Noise signals with equal interfering effects are assigned equal magnitudes. To ac- complish this effect, the American Telephone and Telegraph Company (ATT) developed a weighting network called C-message weighting. When designing the C-message weighting network, groups of observers were asked to adjust the loudness of 14 different frequencies between 180 Hz and 3500 Hz until the sound of each tone was judged to be equally annoying as a 1000-Hz reference tone in the absence of speech. A 1000-Hz tone was selected for the reference because empirical data indicated that 1000 Hz is the most annoying frequency (i.e., the best frequency response) Rdc 0.1095 d2 The Telephone Circuit 409
- 415. FIGURE 2 C-message weighting curve to humans. The same people were then asked to adjust the amplitude of the tones in the pres- ence of speech until the effect of noise on articulation (annoyance) was equal to that of the 1000-Hz reference tone. The results of the two experiments were combined, smoothed, and plotted, resulting in the C-message weighting curve shown in Figure 2. A 500-type tele- phone set was used for these tests; therefore, the C-message weighting curve includes the frequency response characteristics of a standard telephone set receiver as well as the hear- ing response of an average listener. The significance of the C-message weighting curve is best illustrated with an example. From Figure 2, it can be seen that a 200-Hz test tone of a given power is 25 dB less disturbing than a 1000-Hz test tone of the same power. Therefore, the C-mes- sage weighting network will introduce 25 dB more loss for 200 Hz than it will for 1000 Hz. When designing the C-message network, it was found that the additive effect of sev- eral noise sources combine on a root-sum-square (RSS) basis. From these design consider- ations, it was determined that a telephone message–channel noise measuring set should be a voltmeter with the following characteristics: Readings should take into consideration that the interfering effect of noise is a func- tion of frequency as well as magnitude. When dissimilar noise signals are present simultaneously, the meter should combine them to properly measure the overall interfering effect. It should have a transient response resembling that of the human ear. For sounds shorter than 200 ms, the human ear does not fully appreciate the true power of the sound. Therefore, noise-measuring sets are designed to give full-power indication only for bursts of noise lasting 200 ms or longer. The Telephone Circuit 410
- 416. FIGURE 3 3-kHz flat response curve When different types of noise cause equal interference as determined in subjective tests, use of the meter should give equal readings. The reference established for performing message-channel noise measurements is 90 dBm (1012 watts). The power level of 90 dBm was selected because, at the time, power levels could not measure levels below 90 dBm and, therefore, it would not be nec- essary to deal with negative values when reading noise levels. Thus, a 1000-Hz tone with a power level of 90 dBm is equal to a noise reading of 0 dBrn. Conversely, a 1000-Hz tone with a power level of 0 dBm is equal to a noise reading of 90 dBrn, and a 1000-Hz tone with a power level of 40 dBm is equal to a noise reading of 50 dBrn. When appropriate, other weighting networks can be substituted for C-message. For example, a 3-kHz flat network is used to measure power density of white noise. This net- work has a nominal low-pass frequency response down 3 dB at 3 kHz and rolls off at 12 dB per octave. A 3-kHz flat network is often used for measuring high levels of low-frequency noise, such as power supply hum. The frequency response for a 3-kHz flat network is shown in Figure 3. 4 UNITS OF POWER MEASUREMENT 4-1 dB and dBm To specify the amplitudes of signals and interference, it is often convenient to define them at some reference point in the system. The amplitudes at any other physical location can then be related to this reference point if the loss or gain between the two points is known. For example, sea level is generally used as the reference point when comparing elevations. By referencing two mountains to sea level, we can compare the two elevations, regardless of where the mountains are located. A mountain peak in Colorado 12,000 feet above sea level is 4000 feet higher than a mountain peak in France 8000 feet above sea level. The Telephone Circuit 411
- 417. The decibel (dB) is the basic yardstick used for making power measurements in com- munications. The unit dB is simply a logarithmic expression representing the ratio of one power level to another and expressed mathematically as (2) where P1 and P2 are power levels at two different points in a transmission system. From Equation 2, it can be seen when P1 P2, the power ratio is 0 dB; when P1 P2, the power ratio in dB is positive; and when P1 P2, the power ratio in dB is negative. In tele- phone and telecommunications circuits, power levels are given in dBm and differences be- tween power levels in dB. Equation 2 is essentially dimensionless since neither power is referenced to a base. The unit dBm is often used to reference the power level at a given point to 1 milliwatt. One milliwatt is the level from which all dBm measurements are referenced. The unit dBm is an indirect measure of absolute power and expressed mathematically as (3) where P is the power at any point in a transmission system. From Equation 3, it can be seen that a power level of 1 mW equates to 0 dBm, power levels above 1 mW have positive dBm values, and power levels less than 1 mW have negative dBm values. Example 1 Determine a. The power levels in dBm for signal levels of 10 mW and 0.5 mW. b. The difference between the two power levels in dB. Solution a. The power levels in dBm are determined by substituting into Equation 3: b. The difference between the two power levels in dB is determined by substituting into Equation 2: or 10 dBm (3 dBm) 13 dB The 10-mW power level is 13 dB higher than a 0.5-mW power level. Experiments indicate that a listener cannot give a reliable estimate of the loudness of a sound but can distinguish the difference in loudness between two sounds. The ear’s sen- sitivity to a change in sound power follows a logarithmic rather than a linear scale, and the dB has become the unit of this change. 4-2 Transmission Level Point, Transmission Level, and Data Level Point Transmission level point (TLP) is defined as the optimum level of a test tone on a channel at some point in a communications system. The numerical value of the TLP does not de- dB 10 log¢ 10 mW 0.5 mW ≤ 13 dB dBm 10 log¢ 0.5 mW 1 mW ≤ 3 dBm dBm 10 log¢ 10 mW 1 mW ≤ 10 dBm dBm 10 log¢ P 1 mW ≤ dB 10 log¢ P1 P2 ≤ The Telephone Circuit 412
- 418. scribe the total signal power present at that point—it merely defines what the ideal level should be. The transmission level (TL) at any point in a transmission system is the ratio in dB of the power of a signal at that point to the power the same signal would be at a 0-dBm transmission level point. For example, a signal at a particular point in a transmission sys- tem measures 13 dBm. Is this good or bad? This could be answered only if it is known what the signal strength should be at that point. TLP does just that. The reference for TLP is 0 dBm. A 15-dBm TLP indicates that, at this specific point in the transmission system, the signal should measure 15 dBm. Therefore, the transmission level for a signal that measures 13 dBm at a 15-dBm point is 2 dB. A 0 TLP is a TLP where the signal power should be 0 dBm. TLP says nothing about the actual signal level itself. Data level point (DLP) is a parameter equivalent to TLP except TLP is used for voice circuits, whereas DLP is used as a reference for data transmission. The DLP is always 13 dB below the voice level for the same point. If the TLP is 15 dBm, the DLP at the same point is 28 dBm. Because a data signal is more sensitive to nonlinear distortion (harmonic and intermodulation distortion), data signals are transmitted at a lower level than voice sig- nals. 4-3 Units of Measurement Common units for signal and noise power measurements in the telephone industry include dBmO, rn, dBrn, dBrnc, dBrn 3-kHz flat, and dBrncO. 4-3-1 dBmO. dBmO is dBm referenced to a zero transmission level point (0 TLP). dBmO is a power measurement adjusted to 0 dBm that indicates what the power would be if the signal were measured at a 0 TLP. dBmO compares the actual signal level at a point with what that signal level should be at that point. For example, a signal measuring 17 dBm at a 16-dBm transmission level point is 1 dBmO (i.e., the signal is 1 dB be- low what it should be, or if it were measured at a 0 TLP, it would measure 1 dBm). 4-3-2 rn (reference noise). rn is the dB value used as the reference for noise read- ings. Reference noise equals 90 dBm or 1 pW (1 1012 W). This value was selected for two reasons: (1) Early noise measuring sets could not accurately measure noise levels lower than 90 dBm, and (2) noise readings are typically higher than 90 dBm, resulting in positive dB readings in respect to reference noise. 4-3-3 dBrn. dBrn is the dB level of noise with respect to reference noise (90 dBm). dBrn is seldom used by itself since it does not specify a weighting. A noise reading of 50 dBm equates to 40 dBrn, which is 40 dB above reference noise (50 [ 90]) 40 dBrn. 4-3-4 dBrnc. dBrnc is similar to dBrn except dBrnc is the dB value of noise with respect to reference noise using C-message weighting. Noise measurements obtained with a C-message filter are meaningful, as they relate the noise measured to the combined fre- quency response of a standard telephone and the human ear. 4-3-5 dBrn 3-kHz flat. dBrn 3-kHz flat noise measurements are noise readings taken with a filter that has a flat frequency response from 30 Hz to 3 kHz. Noise readings taken with a 3-kHz flat filter are especially useful for detecting low-frequency noise, such as power supply hum. dBrn 3-kHz flat readings are typically 1.5 dB higher than dBrnc read- ings for equal noise power levels. 4-3-6 dBrncO. dBrncO is the amount of noise in dBrnc corrected to a 0 TLP. A noise reading of 34 dBrnc at a 7-dBm TLP equates to 27 dBrncO. dBrncO relates noise power readings (dBrnc) to a 0 TLP. This unit establishes a common reference point through- out the transmission system. The Telephone Circuit 413
- 419. dBm TLP S (-42 dBm) N (-74 dBm) (-2 dBmO) S/N = 32 dB S (48 dBrnc) N (16 dBrnc) (16 dB above rn) Reference noise (rn) S/N = 32 dB dBrnc 0 -5 -10 -15 -20 -25 -30 -35 -40 -45 -50 -55 -60 -65 -70 -75 -80 -85 -90 90 85 80 75 70 65 60 55 50 45 40 35 30 25 20 15 10 5 0 FIGURE 4 Figure for Example 2 Example 2 For a signal measurement of 42 dBm, a noise measurement of 16 dBrnc, and a 40-dBm TLP, determine a. Signal level in dBrnc. b. Noise level in dBm. c. Signal level in dBmO. d. signal-to-noise ratio in dB. (For the solutions, refer to Figure 4.) Solution a. The signal level in dBrnc can be read directly from the chart shown in Figure 4 as 48 dBrnc. The signal level in dBrnc can also be computed mathematically as follows: 42 dBm (90 dBrn) 48 dBrnc b. The noise level in dBm can be read directly from the chart shown in Figure 4 as 74 dBm. The noise level in dBm can also be calculated as follows: 90 16 74 dBm c. The signal level in dBmO is simply the difference between the actual signal level in dBm and the TLP or 2 dBmO as shown in Figure 4. The signal level in dBmO can also be computed mathemati- cally as follows: 42 dBm (40 dBm) 2 dBmO The Telephone Circuit 414
- 420. d. The signal-to-noise ratio is simply the difference in the signal power in dBm and the noise power in dBm or the signal level in dBrnc and the noise power in dBrnc as shown in Figure 4 as 32 dB. The signal-to-noise ratio is computed mathematically as 42 dBm (74 dBm) 32 dB or 48 dBrnc 16 dBrnc 32 dB 4-4 Psophometric Noise Weighting Psophometric noise weighting is used primarily in Europe. Psophometric weighting as- sumes a perfect receiver; therefore, its weighting curve corresponds to the frequency re- sponse of the human ear only. The difference between C-message weighting and psopho- metric weighting is so small that the same conversion factor may be used for both. 5 TRANSMISSION PARAMETERS AND PRIVATE-LINE CIRCUITS Transmission parameters apply to dedicated private-line data circuits that utilize the private sector of the public telephone network—circuits with bandwidths comparable to those of standard voice-grade telephone channels that do not utilize the public switched telephone network. Private-line circuits are direct connections between two or more lo- cations. On private-line circuits, transmission facilities and other telephone company– provided equipment are hardwired and available only to a specific subscriber. Most private-line data circuits use four-wire, full-duplex facilities. Signal paths established through switched lines are inconsistent and may differ greatly from one call to an- other. In addition, telephone lines provided through the public switched telephone net- work are two wire, which limits high-speed data transmission to half-duplex opera- tion. Private-line data circuits have several advantages over using the switched public telephone network: Transmission characteristics are more consistent because the same facilities are used with every transmission. The facilities are less prone to noise produced in telephone company switches. Line conditioning is available only on private-line facilities. Higher transmission bit rates and better performance is appreciated with private-line data circuits. Private-line data circuits are more economical for high-volume circuits. Transmission parameters are divided into three broad categories: bandwidth pa- rameters, which include attenuation distortion and envelope delay distortion; interface parameters, which include terminal impedance, in-band and out-of-band signal power, test signal power, and ground isolation; and facility parameters, which include noise measurements, frequency distortion, phase distortion, amplitude distortion, and non- linear distortion. 5-1 Bandwidth Parameters The only transmission parameters with limits specified by the FCC are attenuation distor- tion and envelope delay distortion. Attenuation distortion is the difference in circuit gain experienced at a particular frequency with respect to the circuit gain of a reference fre- quency. This characteristic is sometimes referred to as frequency response, differential gain, and 1004-Hz deviation. Envelope delay distortion is an indirect method of evaluating the phase delay characteristics of a circuit. FCC tariffs specify the limits for attenuation dis- tortion and envelope delay distortion. To reduce attenuation and envelope delay distortion The Telephone Circuit 415
- 421. and improve the performance of data modems operating over standard message channels, it is often necessary to improve the quality of the channel. The process used to improve a basic telephone channel is called line conditioning. Line conditioning improves the high- frequency response of a message channel and reduces power loss. The attenuation and delay characteristics of a circuit are artificially altered to meet limits prescribed by the line conditioning requirements. Line conditioning is available only to private-line subscribers at an additional charge. The basic voice-band channel (some- times called a basic 3002 channel) satisfies the minimum line conditioning requirements. Telephone companies offer two types of special line conditioning for subscriber loops: C-type and D-type. 5-1-1 C-type line conditioning. C-type conditioning specifies the maximum limits for attenuation distortion and envelope delay distortion. C-type conditioning per- tains to line impairments for which compensation can be made with filters and equal- izers. This is accomplished with telephone company–provided equipment. When a cir- cuit is initially turned up for service with a specific C-type conditioning, it must meet the requirements for that type of conditioning. The subscriber may include devices within the station equipment that compensate for minor long-term variations in the bandwidth requirements. There are five classifications or levels of C-type conditioning available. The grade of conditioning a subscriber selects depends on the bit rate, modulation technique, and desired performance of the data modems used on the line. The five classifications of C-type condi- tioning are the following: C1 and C2 conditioning pertain to two-point and multipoint circuits. C3 conditioning is for access lines and trunk circuits associated with private switched networks. C4 conditioning pertains to two-point and multipoint circuits with a maximum of four stations. C5 conditioning pertains only to two-point circuits. Private switched networks are telephone systems provided by local telephone com- panies dedicated to a single customer, usually with a large number of stations. An example is a large corporation with offices and complexes at two or more geographical locations, sometimes separated by great distances. Each location generally has an on-premise private branch exchange (PBX). A PBX is a relatively low-capacity switching machine where the subscribers are generally limited to stations within the same building or building com- plex. Common-usage access lines and trunk circuits are required to interconnect two or more PBXs. They are common only to the subscribers of the private network and not to the general public telephone network. Table 1 lists the limits prescribed by C-type con- ditioning for attenuation distortion. As the table shows, the higher the classification of conditioning imposed on a circuit, the flatter the frequency response and, therefore, a bet- ter-quality circuit. Attenuation distortion is simply the frequency response of a transmission medium referenced to a 1004-Hz test tone. The attenuation for voice-band frequencies on a typical cable pair is directly proportional to the square root of the frequency. From Table 1, the at- tenuation distortion limits for a basic (unconditioned) circuit specify the circuit gain at any frequency between 500 Hz and 2500 Hz to be not more than 2 dB more than the circuit gain at 1004 Hz and not more than 3 dB below the circuit gain at 1004 Hz. For attenuation dis- tortion, the circuit gain for 1004 Hz is always the reference. Also, within the frequency bands from 300 Hz and 499 Hz and from 2501 Hz to 3000 Hz, the circuit gain cannot be The Telephone Circuit 416
- 422. Basic and C-Type Conditioning Requirements Attenuation Distortion Envelope Delay (Frequency Response Relative to 1004 Hz) Distortion Channel Frequency Variation Frequency Variation Conditioning Range (Hz) (dB) Range (Hz) (μs) Basic 300–499 3 to 12 800–2600 1750 500–2500 2 to 8 2501–3000 3 to 12 C1 300–999 2 to 6 800–999 1750 1000–2400 1 to 3 1000–2400 1000 2401–2700 3 to 6 2401–2600 1750 2701–3000 3 to 12 C2 300–499 2 to 6 500–600 3000 500–2800 1 to 3 601–999 1500 2801–3000 2 to 6 1000–2600 500 2601–2800 3000 C3 (access line) 300–499 0.8 to 3 500–599 650 500–2800 0.5 to 1.5 600–999 300 2801–3000 0.8 to 3 1000–2600 110 2601–2800 650 C3 (trunk) 300–499 0.8 to 2 500–599 500 500–2800 0.5 to 1 600–999 260 2801–3000 0.8 to 2 1000–2600 80 2601–3000 500 C4 300–499 2 to 6 500–599 3000 500–3000 2 to 3 600–799 1500 3001–3200 2 to 6 800–999 500 1000–2600 300 2601–2800 500 2801–3000 1500 C5 300–499 1 to 3 500–599 600 500–2800 0.5 to 1.5 600–999 300 2801–3000 1 to 3 1000–2600 100 2601–2800 600 greater than 3 dB above or more than 12 dB below the gain at 1004 Hz. Figure 5 shows a graphical presentation of basic line conditioning requirements. Figure 6 shows a graphical presentation of the attenuation distortion requirements spec- ified in Table 1 for C2 conditioning, and Figure 7 shows the graph for C2 conditioning super- imposed over the graph for basic conditioning. From Figure 7, it can be seen that the require- ments for C2 conditioning are much more stringent than those for a basic circuit. Example 3 A 1004 Hz test tone is transmitted over a telephone circuit at 0 dBm and received at 16 dBm. Determine a. The 1004-Hz circuit gain. b. The attenuation distortion requirements for a basic circuit. c. The attenuation distortion requirements for a C2 conditioned circuit. Solution a. The circuit gain is determined mathematically as 0 dBm (16 dB) 16 dB (which equates to a loss of 16 dB) The Telephone Circuit Table 1 417
- 423. FIGURE 5 Graphical presentation of the limits for attenuation distortion for a basic 3002 telephone circuit FIGURE 6 Graphical presentation of the limits for attenuation distortion for a C2 conditioned telephone circuit b. Circuit gain requirements for a basic circuit can be determined from Table 1: Frequency Band Requirements Minimum Level Maximum Level 500 Hz and 2500 Hz 2 dB and 8 dB 24 dBm 14 dBm 300 Hz and 499 Hz 3 dB and 12 dB 28 dBm 13 dBm 2501 Hz and 3000 Hz 3 dB and 12 dB 28 dBm 13 dBm The Telephone Circuit 418
- 424. FIGURE 7 Overlay of Figure 5 over Figure 6 to demonstrate the more stringent requirements imposed by C2 conditioning compared to a basic (unconditioned) circuit c. Circuit gain requirements for a C2 conditioned circuit can be determined from Table 1: Frequency Band Requirements Minimum Level Maximum Level 500 Hz and 2500 Hz 1 dB and 3 dB 19 dBm 15 dBm 300 Hz and 499 Hz 2 dB and 6 dB 22 dBm 14 dBm 2801 Hz and 3000 Hz 2 dB and 6 dB 22 dBm 14 dBm A linear phase-versus-frequency relationship is a requirement for error-free data transmission—signals are delayed more at some frequencies than others. Delay distortion is the difference in phase shifts with respect to frequency that signals experience as they propagate through a transmission medium. This relationship is difficult to measure because of the difficulty in establishing a phase (time) reference. Envelope delay is an alternate method of evaluating the phase-versus-frequency relationship of a circuit. The time delay encountered by a signal as it propagates from a source to a destina- tion is called propagation time, and the delay measured in angular units, such as degrees or radians, is called phase delay. All frequencies in the usable voice band (300 Hz to 3000 Hz) do not experience the same time delay in a circuit. Therefore, a complex waveform, such as the output of a data modem, does not possess the same phase-versus-frequency relationship when received as it possessed when it was transmitted. This condition represents a possible impairment to a data signal. The absolute phase delay is the actual time required for a par- ticular frequency to propagate from a source to a destination through a communications channel. The difference between the absolute delays of all the frequencies is phase distor- tion. A graph of phase delay-versus-frequency for a typical circuit is nonlinear. By definition, envelope delay is the first derivative (slope) of phase with respect to frequency: (4) envelope delay dθ1ω2 dω The Telephone Circuit 419
- 425. FIGURE 8 Graphical presentation of the limits for envelope delay in a basic telephone channel In actuality, envelope delay only closely approximates dθ(ω)/dω. Envelope delay mea- surements evaluate not the true phase-versus-frequency characteristics but rather the phase of a wave that is the result of a narrow band of frequencies. It is a common misconception to confuse true phase distortion (also called delay distortion) with envelope delay distortion (EDD). Envelope delay is the time required to propagate a change in an AM envelope (the actual information-bearing part of the signal) through a transmission medium. To measure envelope delay, a narrowband amplitude-modulated carrier, whose frequency is varied over the usable voice band, is transmitted (the amplitude-modulated rate is typically between 25 Hz and 100 Hz). At the receiver, phase variations of the low-frequency envelope are measured. The phase difference at the different carrier frequencies is envelope delay dis- tortion. The carrier frequency that produces the minimum envelope delay is established as the reference and is normalized to zero. Therefore, EDD measurements are typically given in microseconds and yield only positive values. EDD indicates the relative envelope delays of the various carrier frequencies with respect to the reference frequency. The reference fre- quency of a typical voice-band circuit is typically around 1800 Hz. EDD measurements do not yield true phase delays, nor do they determine the relative relationships between true phase delays. EDD measurements are used to determine a close approximation of the relative phase delay characteristics of a circuit. Propagation time can- not be increased. Therefore, to correct delay distortion, equalizers are placed in a circuit to slow down the frequencies that travel the fastest more than frequencies that travel the slow- est. This reduces the difference between the fastest and slowest frequencies, reducing the phase distortion. The EDD limits for basic and conditioned telephone channels are listed in Table 1. Figure 8 shows a graphical representation of the EDD limits for a basic telephone channel, The Telephone Circuit 420
- 426. FIGURE 9 Graphical presentation of the limits for envelope delay in a telephone channel with C2 conditioning and Figure 9 shows a graphical representation of the EDD limits for a channel meeting the requirements for C2 conditioning. From Table 1, the EDD limit of a basic telephone chan- nel is 1750 μs between 800 Hz and 2600 Hz. This indicates that the maximum difference in envelope delay between any two carrier frequencies (the fastest and slowest frequencies) within this range cannot exceed 1750 μs. Example 4 An EDD test on a basic telephone channel indicated that an 1800-Hz carrier experienced the mini- mum absolute delay of 400 μs. Therefore, it is the reference frequency. Determine the maximum ab- solute envelope delay that any frequency within the 800-Hz to 2600-Hz range can experience. Solution The maximum envelope delay for a basic telephone channel is 1750 μs within the frequency range of 800 Hz to 2600 Hz. Therefore, the maximum envelope delay is 2150 μs (400 μs 1750 μs). The absolute time delay encountered by a signal between any two points in the con- tinental United States should never exceed 100 ms, which is not sufficient to cause any problems. Consequently, relative rather than absolute values of envelope delay are mea- sured. For the previous example, as long as EDD tests yield relative values less than 1750 μs, the circuit is within limits. 5-1-2 D-type line conditioning. D-type conditioning neither reduces the noise on a circuit nor improves the signal-to-noise ratio. It simply sets the minimum requirements for signal-to-noise (S/N) ratio and nonlinear distortion. If a subscriber requests D-type conditioning and the facilities assigned to the circuit do not meet the requirements, a dif- ferent facility is assigned. D-type conditioning is simply a requirement and does not add The Telephone Circuit 421
- 427. anything to the circuit, and it cannot be used to improve a circuit. It simply places higher requirements on circuits used for high-speed data transmission. Only circuits that meet D-type conditioning requirements can be used for high-speed data transmission. D-type conditioning is sometimes referred to as high-performance conditioning and can be applied to private-line data circuits in addition to either basic or C-conditioned requirements. There are two categories for D-type conditioning: D1 and D2. Limits imposed by D1 and D2 are virtually identical. The only difference between the two categories is the circuit arrange- ment to which they apply. D1 conditioning specifies requirements for two-point circuits, and D2 conditioning specifies requirements for multipoint circuits. D-type conditioning is mandatory when the data transmission rate is 9600 bps be- cause without D-type conditioning, it is highly unlikely that the circuit can meet the mini- mum performance requirements guaranteed by the telephone company. When a telephone company assigns a circuit to a subscriber for use as a 9600-bps data circuit and the circuit does not meet the minimum requirements of D-type conditioning, a new circuit is assigned. This is because a circuit cannot generally be upgraded to meet D-type conditioning speci- fications by simply adding corrective devices, such as equalizers and amplifiers. Telephone companies do not guarantee the performance of data modems operating at bit rates above 9600 bps over standard voice-grade circuits. D-type conditioned circuits must meet the following specifications: Signal-to-C-notched noise ratio: 28 dB Nonlinear distortion Signal-to-second order distortion: 35 dB Signal-to-third order distortion: 40 dB The signal-to-notched noise ratio requirement for standard circuits is only 24 dB, and they have no requirements for nonlinear distortion. Nonlinear distortion is an example of correlated noise and is produced from non- linear amplification. When an amplifier is driven into a nonlinear operating region, the signal is distorted, producing multiples and sums and differences (cross products) the original signal frequencies. The noise caused by nonlinear distortion is in the form of additional frequencies produced from nonlinear amplification of a signal. In other words, no signal, no noise. Nonlinear distortion produces distorted waveforms that are detrimental to digitally modulated carriers used with voice-band data modems, such as FSK, PSK, and QAM. Two classifications of nonlinear distortion are harmonic distor- tion (unwanted multiples of the transmitted frequencies) and intermodulation distor- tion (cross products [sums and differences] of the transmitted frequencies, sometimes called fluctuation noise or cross-modulation noise). Harmonic and intermodulation dis- tortion, if of sufficient magnitude, can destroy the integrity of a data signal. The degree of circuit nonlinearity can be measured using either harmonic or intermodulation dis- tortion tests. Harmonic distortion is measured by applying a single-frequency test tone to a tele- phone channel. At the receive end, the power of the fundamental, second, and third har- monic frequencies is measured. Harmonic distortion is classified as second, third, nth or- der, or as total harmonic distortion. The actual amount of nonlinearity in a circuit is determined by comparing the power of the fundamental with the combined powers of the second and third harmonics. Harmonic distortion tests use a single-frequency (704-Hz) source (see Figure 10); therefore, no cross-product frequencies are produced. Although simple harmonic distortion tests provide an accurate measurement of the nonlinear characteristics of analog telephone channel, they are inadequate for digital (T car- rier) facilities. For this reason, a more refined method was developed that uses a multifre- quency test-tone signal. Four test frequencies are used (see Figure 11): two designated The Telephone Circuit 422
- 428. V1 704 Hz x 100 = 2nd order harmonic distortion 704 Hz V2 V3 f1 = 704 Hz Fundamental V2 V1 f2 = 2f1 = 1408 Hz 2nd Harmonic f3 = 3f1 = 2112 Hz 3rd Harmonic x 100 = 3rd order harmonic distortion x 100 = total harmonic distortion (THD) V3 V1 V1 V2 2 + V3 2 √ FIGURE 10 Harmonic distortion 4-Tone test signal A-Band 2nd order 3rd order 2nd order B-Band A1 A2 A1 = 856 Hz A2 = 863 Hz B1 = 1374 Hz B2 = 1385 Hz B – A 2B – A B + A B1 B2 FIGURE 11 Intermodulation distortion the A band (A1 856 Hz, A2 863 Hz) and two designated the B band (B1 1374 Hz and B2 1385 Hz). The four frequencies are transmitted with equal power levels, and the total combined power is equal to that of a normal data signal. The nonlinear amplification of the circuit produces multiples of each frequency (harmonics) and their cross-product fre- quencies (sum and difference frequencies). For reasons beyond the scope of this text, the following second- and third-order products were selected for measurement: B A, B A, and 2B A. The combined signal power of the four A and B band frequencies is compared with the second-order cross products and then compared with the third-order cross prod- ucts. The results are converted to dB values and then compared to the requirements of D-type conditioning. Harmonic and intermodulation distortion tests do not directly determine the amount of interference caused by nonlinear circuit gain. They serve as a figure of merit only when evaluating circuit parameters. 5-2 Interface Parameters The two primary considerations of the interface parameters are electrical protection of the telephone network and its personnel and standardization of design arrangements. The in- terface parameters include the following: Station equipment impedances should be 600 Ω resistive over the usable voice band. Station equipment should be isolated from ground by a minimum of 20 MΩ dc and 50 kΩ ac. The basic voice-grade telephone circuit is a 3002 channel; it has an ideal bandwidth of 0 Hz to 4 kHz and a usable bandwidth of 300 Hz to 3000 Hz. The circuit gain at 3000 Hz is 3 dB below the specified in-band signal power. The gain at 4 kHz must be at least 15 dB below the gain at 3 kHz. The maximum transmitted signal power for a private-line circuit is 0 dBm. The transmitted signal power for dial-up circuits using the public switched telephone network is established for each loop so that the signal is received at the telephone cen- tral office at 12 dBm. Table 2 summarizes interface parameter limits. The Telephone Circuit 423
- 429. Interface Parameter Limits Parameter Limit 1. Recommended impedance of terminal equipment 600 Ω resistive 10% 2. Recommended isolation to ground of terminal At least 20 MΩ dc equipment At least 50 kΩ ac At least 1500 V rms breakdown voltage at 60 Hz 3. Data transmit signal power 0 dBm (3-s average) 4. In-band transmitted signal power 2450-Hz to 2750-Hz band should not exceed signal power in 800-Hz to 2450-Hz band 5. Out-of-band transmitted signal power Above voice band: (a) 3995 Hz–4005 Hz At least 18 dB below maximum allowed in-band signal power (b) 4-kHz–10-kHz band Less than –16 dBm (c) 10-kHz–25-kHz band Less than –24 dBm (d) 25-kHz–40-kHz band Less than –36 dBm (e) Above 40 kHz Less than –50 dBm Below voice band: (f) rms current per conductor as specified by Telco but never greater than 0.35 A. (g) Magnitude of peak conductor-to-ground voltage not to exceed 70 V. (h) Conductor-to-conductor voltage shall be such that conductor-to-ground voltage is not exceeded. For an underground signal source, the conductor-to-conductor limit is the same as the conductor-to-ground limit. (i) Total weighted rms voltage in band from 50 Hz to 300 Hz, not to exceed 100 V. Weighting factors for each frequency component (f) are f2 /104 for f between 50 Hz and 100 Hz and f3.3 /106.6 for f between 101 Hz and 300 Hz. 6. Maximum test signal power: same as transmitted data power. 5-3 Facility Parameters Facility parameters represent potential impairments to a data signal. These impairments are caused by telephone company equipment and the limits specified pertain to all private-line data circuits using voice-band facilities, regardless of line conditioning. Facility parameters include 1004-Hz variation, C-message noise, impulse noise, gain hits and dropouts, phase hits, phase jitter, single-frequency interference, frequency shift, phase intercept distortion, and peak-to-average ratio. 5-3-1 1004-Hz variation. The telephone industry has established 1004 Hz as the stan- dard test-tone frequency; 1000 Hz was originally selected because of its relative location in the passband of a standard voice-band circuit.The frequency was changed to 1004 Hz with the ad- vent of digital carriers because 1000 Hz is an exact submultiple of the 8-kHz sample rate used with T carriers. Sampling a continuous 1000-Hz signal at an 8000-Hz rate produced repetitive patterns in the PCM codes, which could cause the system to lose frame synchronization. The purpose of the 1004-Hz test tone is to simulate the combined signal power of a standard voice-band data transmission. The 1004-Hz channel loss for a private-line data cir- cuit is typically 16 dB. A 1004-Hz test tone applied at the transmit end of a circuit should be received at the output of the circuit at 16 dBm. Long-term variations in the gain of the transmission facility are called 1004-Hz variation and should not exceed 4 dB. Thus, the received signal power must be within the limits of 12 dBm to 20 dBm. 5-3-2 C-message noise. C-message noise measurements determine the average weighted rms noise power. Unwanted electrical signals are produced from the random movement of electrons in conductors. This type of noise is commonly called thermal noise because its magnitude is directly proportional to temperature. Because the electron movement is completely random and travels in all directions, thermal noise is also called random noise, and because it contains all frequencies, it is sometimes referred to as white noise. Thermal The Telephone Circuit Table 2 424
- 430. Telephone or data circuit Telephone company facilities Subscriber's location 600-Ω termination Subscriber's location C-message filter Noise power meter Telephone set or data modem Telephone set or data modem FIGURE 12 Terminated C-message noise test setup Telephone or data circuit Telephone company facilities Subscriber's location Subscriber's location C- notched filter Telephone set or data modem 1004-Hz oscillator Noise power meter Telephone set or data modem C- message filter FIGURE 13 C-notched noise test setup noise is inherently present in a circuit because of its electrical makeup. Because thermal noise is additive, its magnitude is dependent, in part, on the electrical length of the circuit. C-messagenoisemeasurementsaretheterminatedrmspowerreadingsatthereceiveend of a circuit with the transmit end terminated in the characteristic impedance of the telephone line. Figure 12 shows the test setup for conducting terminated C-message noise readings. As shown in the figure, a C-message filter is placed between the circuit and the power meter in the noise measuring set so that the noise measurement evaluates the noise with a response similar to that of a human listening to the noise through a standard telephone set speaker. There is a disadvantage to measuring noise this way. The overall circuit characteris- tics, in the absence of a signal, are not necessarily the same as when a signal is present. Us- ing compressors, expanders, and automatic gain devices in a circuit causes this difference. For this reason, C-notched noise measurements were developed. C-notched noise mea- surements differ from standard C-message noise measurements only in the fact that a holding tone (usually 1004 Hz or 2804 Hz) is applied to the transmit end of the circuit while the noise measurement is taken. The holding tone ensures that the circuit operation simu- lates a loaded voice or data transmission. Loaded is a communications term that indicates the presence of a signal power comparable to the power of an actual message transmission. A narrowband notch filter removes the holding tone before the noise power is measured. The test setup for making C-notched noise measurements is shown in Figure 13. As the fig- ure shows, the notch filter is placed in front of the C-message filter, thus blocking the hold- ing tone from reaching the power meter. The Telephone Circuit 425
- 431. 16 dB 0 dBm 0 TLP -16 dBm Voice signal level -29 dBm Data signal level -35 dBm Impulse noise threshold Thermal noise threshold level Thermal noise 24 dB minimum C-notched noise-to-signal ratio 28 dB minimum C-notched noise-to-signal ratio Impulse noise hit -53 dBm Standard circuit -57 dBm High performance (D-conditioning) 13 dB 6 dB FIGURE 14 C-notched noise and impulse noise The physical makeup of a private-line data circuit may require using several carrier facilities and cable arrangements in tandem. Each facility may be analog, digital, or some combination of analog and digital. Telephone companies have established realistic C-notched noise requirements for each type of facility for various circuit lengths. Telephone compa- nies guarantee standard private-line data circuits a minimum signal-to-C-notched noise ra- tio of 24 dB. A standard circuit is one operating at less than 9600 bps. Data circuits operat- ing at 9600 bps require D-type conditioning, which guarantees a minimum signal-to- C-notched noise ratio of 28 dB. C-notched noise is shown in Figure 14. Telephone compa- nies do not guarantee the performance of voice-band circuits operating at bit rates in excess of 9600 bps. 5-3-3 Impulse noise. Impulse noise is characterized by high-amplitude peaks (im- pulses) of short duration having an approximately flat frequency spectrum. Impulse noise can saturate a message channel. Impulse noise is the primary source of transmission errors in data circuits. There are numerous sources of impulse noise—some are controllable, but most are not. The primary cause of impulse noise is man-made sources, such as interference from ac powerlines,transientsfromswitchingmachines,motors,solenoids,relays,electrictrains,and so on. Impulse noise can also result from lightning and other adverse atmospheric conditions. The significance of impulse noise hits on data transmission has been a controversial topic. Telephone companies have accepted the fact that the absolute magnitude of the im- pulse hit is not as significant as the magnitude of the hit relative to the signal amplitude. Empirically, it has been determined that an impulse hit will not produce transmission errors in a data signal unless it comes within 6 dB of the signal level as shown in Figure 14. Im- pulse hit counters are designed to register a maximum of seven counts per second. This leaves a 143-ms lapse called a dead time between counts when additional impulse hits are not registered. Contemporary high-speed data formats transfer data in a block or frame for- mat, and whether one hit or many hits occur during a single transmission is unimportant, as any error within a message generally necessitates retransmission of the entire message. It has been determined that counting additional impulses during the time of a single trans- mission does not correlate well with data transmission performance. The Telephone Circuit 426
- 432. The Telephone Circuit 4 ms Positive gain hit Negative gain hit Dropout Signal reference level +3 dB 0 dB -3 dB -12 dB 4 ms 4 ms FIGURE 15 Gain hits and dropouts Impulse noise objectives are based primarily on the error susceptibility of data sig- nals, which depends on the type of modem used and the characteristics of the transmission medium. It is impractical to measure the exact peak amplitudes of each noise pulse or to count the number that occur. Studies have shown that expected error rates in the absence of other impairments are approximately proportional to the number of impulse hits that ex- ceed the rms signal power level by approximately 2 dB. When impulse noise tests are per- formed, a 2802-Hz holding tone is placed on a circuit to ensure loaded circuit conditions. The counter records the number of hits in a prescribed time interval (usually 15 minutes). An impulse hit is typically less than 4 ms in duration and never more than 10 ms. Telephone company limits for recordable impulse hits is 15 hits within a 15-minute time interval. This does not limit the number of hits to one per minute but, rather, the average occurrence to one per minute. 5-3-4 Gain hits and dropouts. A gain hit is a sudden, random change in the gain of a circuit resulting in a temporary change in the signal level. Gain hits are classified as temporary variations in circuit gain exceeding 3 dB, lasting more than 4 ms, and return- ing to the original value within 200 ms. The primary cause of gain hits is noise transients (impulses) on transmission facilities during the normal course of a day. A dropout is a decrease in circuit gain (i.e., signal level) of more than 12 dB lasting longer than 4 ms. Dropouts are characteristics of temporary open-circuit conditions and are generally caused by deep fades on radio facilities or by switching delays. Gain hits and dropouts are depicted in Figure 15. 5-3-5 Phase hits. Phase hits (slips) are sudden, random changes in the phase of a signal. Phase hits are classified as temporary variations in the phase of a signal lasting longer than 4 ms. Generally, phase hits are not recorded unless they exceed 20C° peak. Phase hits, like gain hits, are caused by transients produced when transmission facilities are switched. Phase hits are shown in Figure 16. 427
- 433. The Telephone Circuit Telephone set Modem Communications channel Communications channel Input frequency spectrum Communications channel Output frequency spectrum Spurious tone 3 6 9 # 2 5 8 0 1 4 7 * Telephone set Modem 3 6 9 # 2 5 8 0 1 4 7 * FIGURE 17 Single-frequency interference (spurious tone) 5-3-6 Phase jitter. Phase jitter is a form of incidental phase modulation—a contin- uous, uncontrolled variation in the zero crossings of a signal. Generally, phase jitter occurs at a 300-Hz rate or lower, and its primary cause is low-frequency ac ripple in power sup- plies. The number of power supplies required in a circuit is directly proportional to the num- ber of transmission facilities and telephone offices that make up the message channel. Each facility has a separate phase jitter requirement; however, the maximum acceptable end-to- end phase jitter is 10° peak to peak regardless of how many transmission facilities or tele- phone offices are used in the circuit. Phase jitter is shown in Figure 16. 5-3-7 Single-frequency interference. Single-frequency interference is the presence of one or more continuous, unwanted tones within a message channel. The tones are called spurious tones and are often caused by crosstalk or cross modulation between adjacent channels in a transmission system due to system nonlinearities. Spurious tones are meas- ured by terminating the transmit end of a circuit and then observing the channel frequency band. Spurious tones can cause the same undesired circuit behavior as thermal noise. Sin- gle-frequency interference is shown in Figure 17. 5-3-8 Frequency shift. Frequency shift is when the frequency of a signal changes dur- ing transmission. For example, a tone transmitted at 1004 Hz is received at 1005 Hz. Analog transmission systems used by telephone companies operate single-sideband suppressed carrier (SSBSC) and, therefore, require coherent demodulation.With coherent demodulation, carriers must be synchronous—the frequency must be reproduced exactly in the receiver. If this is not accomplished, the demodulated signal will be offset in frequency by the difference between transmitandreceivecarrierfrequencies.Thelongeracircuit,themoreanalogtransmissionsys- tems and the more likely frequency shift will occur. Frequency shift is shown in Figure 18. FIGURE 16 Phase hits and phase jitter 428
- 434. The Telephone Circuit 100,000 Hz carrier oscillator 99,998-Hz carrier oscillator Modulator Input 1004 Hz test signal Input 1004 Hz Output difference frequency (101,004 Hz – 99,998 Hz) 101,004 Hz sum frequency (100,000 Hz + 1004 Hz) Output 1006 Hz Demodulator Communications channel FIGURE 18 Frequency shift 5-3-9 Phase intercept distortion. Phase intercept distortion occurs in coherent SSBSC systems, such as those using frequency-division multiplexing when the received carrier is not reinserted with the exact phase relationship to the received signal as the trans- mit carrier possessed. This impairment causes a constant phase shift to all frequencies, which is of little concern for data modems using FSK, PSK, or QAM. Because these are practically the only techniques used today with voice-band data modems, no limits have been set for phase intercept distortion. 5-3-10 Peak-to-average ratio. The difficulties encountered in measuring true phase distortion or envelope delay distortion led to the development of peak-to-average ra- tio (PAR) tests. A signal containing a series of distinctly shaped pulses with a high peak voltage-to-average voltage ratio is transmitted. Differential delay distortion in a circuit has a tendency to spread the pulses, thus reducing the peak voltage-to-average voltage ratio. Low peak-to-average ratios indicate the presence of differential delay distortion. PAR measurements are less sensitive to attenuation distortion than EDD tests and are easier to perform. 5-3-11 Facility parameter summary. Table 3 summarizes facility parameter lim- its. 6 VOICE-FREQUENCY CIRCUIT ARRANGEMENTS Electronic communications circuits can be configured in several ways. Telephone instru- ments and the voice-frequency facilities to which they are connected may be either two wire or four wire. Two-wire circuits have an obvious economic advantage, as they use only half as much copper wire. This is why most local subscriber loops connected to the public switched telephone network are two wire. However, most private-line data circuits are con- figured four wire. 6-1 Two-Wire Voice-Frequency Circuits As the name implies, two-wire transmission involves two wires (one for the signal and one for a reference or ground) or a circuit configuration that is equivalent to using only two wires. Two-wire circuits are ideally suited to simplex transmission, although they are often used for half- and full-duplex transmission. Figure 19 shows the block diagrams for four possible two-wire circuit configura- tions. Figure 19a shows the simplest two-wire configuration, which is a passive circuit consisting of two copper wires connecting a telephone or voice-band modem at one sta- tion through a telephone company interface to a telephone or voice-band modem at the 429
- 435. The Telephone Circuit destination station. The modem, telephone, and circuit configuration are capable of two- way transmission in either the half- or the full-duplex mode. Figure 19b shows an active two-wire transmission system (i.e., one that provides gain). The only difference between this circuit and the one shown in Figure 19a is the ad- dition of an amplifier to compensate for transmission line losses. The amplifier is unidirec- tional and, thus, limits transmission to one direction only (simplex). Figure 19c shows a two-wire circuit using a digital T carrier for the transmission medium. This circuit requires a T carrier transmitter at one end and a T carrier receiver at the other end. The digital T carrier transmission line is capable of two-way transmission; however, the transmitter and receiver in the T carrier are not. The transmitter encodes the analog voice or modem signals into a PCM code, and the decoder in the receiver performs the opposite operation, converting PCM codes back to analog. The digital transmission medium is a pair of copper wire. Figures 19a, b, and c are examples of physical two-wire circuits, as the two stations are physically interconnected with a two-wire metallic transmission line. Figure 19d shows an equivalent two-wire circuit. The transmission medium is Earth’s atmosphere, and there Table 3 Facility Parameter Limits Parameter Limit 1. 1004-Hz loss variation Not more than 4 dB long term 2. C-message noise Maximum rms noise at modem receiver (nominal 16 dBm point) Facility miles dBm dBrncO 0–50 61 32 51–100 59 34 101–400 58 35 401–1000 55 38 1001–1500 54 39 1501–2500 52 41 2501–4000 50 43 4001–8000 47 46 8001–16,000 44 49 3. C-notched noise (minimum values) (a) Standard voice-band channel 24-dB signal to C-notched noise (b) High-performance line 28-dB signal to C-notched noise 4. Single-frequency interference At least 3 dB below C-message noise limits 5. Impulse noise Threshold with respect to Maximum counts above threshold 1004-Hz holding tone allowed in 15 minutes 0 dB 15 4 dB 9 8 dB 5 6. Frequency shift 5 Hz end to end 7. Phase intercept distortion No limits 8. Phase jitter No more than 10° peak to peak (end-to-end requirement) 9. Nonlinear distortion (D-conditioned circuits only) Signal to second order At least 35 dB Signal to third order At least 40 dB 10. Peak-to-average ratio Reading of 50 minimum end to end with standard PAR meter 11. Phase hits 8 or less in any 15-minute period greater than 20 peak 12. Gain hits 8 or less in any 15-minute period greater than 3 dB 13. Dropouts 2 or less in any 15-minute period greater than 12 dB 430
- 436. The Telephone Circuit Telco Interface Tx/Rx Telco Interface Tx/Rx Passive two-wire transmission line Two-wire Two-wire Station B Station A (a) Bidirectional 3 6 9 # 2 5 8 0 1 4 7 * 3 6 9 # 2 5 8 0 1 4 7 * Two-wire Two-wire Telco Interface Tx Telco Interface Rx Active two-wire transmission line Two-wire Two-wire Station B Station A (b) Unidirectional 3 6 9 # 2 5 8 0 1 4 7 * 3 6 9 # 2 5 8 0 1 4 7 * Two-wire Two-wire Amp FIGURE 19 Two-wire configurations: (a) passive cable circuit; (b) active cable circuit (Continued) are no copper wires between the two stations. Although Earth’s atmosphere is capable of two-way simultaneous transmission, the radio transmitter and receiver are not. Therefore, this is considered an equivalent two-wire circuit. 6-2 Four-Wire Voice-Frequency Circuits As the name implies, four-wire transmission involves four wires (two for each direction— a signal and a reference) or a circuit configuration that is equivalent to using four wires. Four-wire circuits are ideally suited to full-duplex transmission, although they can (and very often do) operate in the half-duplex mode. As with two-wire transmission, there are two forms of four-wire transmission systems: physical four wire and equivalent four wire. Figure 20 shows the block diagrams for four possible four-wire circuit configura- tions. As the figures show, a four-wire circuit is equivalent to two two-wire circuits, one for each direction of transmission. The circuits shown in Figures 20a, b, and c are physical four- wire circuits, as the transmitter at one station is hardwired to the receiver at the other sta- tion. Therefore, each two-wire pair is unidirectional (simplex), but the combined four-wire circuit is bidirectional (full duplex). The circuit shown in Figure 20d is an equivalent four-wire circuit that uses Earth’s at- mosphere for the transmission medium. Station A transmits on one frequency (f1) and re- ceives on a different frequency (f2), while station B transmits on frequency f2 and receives on frequency f1. Therefore, the two radio signals do not interfere with one another, and si- multaneous bidirectional transmission is possible. 431
- 437. The Telephone Circuit 6-3 Two Wire versus Four Wire There are several inherent advantages of four-wire circuits over two-wire circuits. For in- stance, four-wire circuits are considerably less noisy, have less crosstalk, and provide more isolation between the two directions of transmission when operating in either the half- or the full-duplex mode. However, two-wire circuits require less wire, less circuitry and, thus, less money than their four-wire counterparts. Providing amplification is another disadvantage of four-wire operation. Telephone or modem signals propagated more than a few miles require amplification. A bidirec- tional amplifier on a two-wire circuit is not practical. It is much easier to separate the two directions of propagation with a four-wire circuit and install separate amplifiers in each direction. 6-4 Hybrids, Echo Suppressors, and Echo Cancelers When a two-wire circuit is connected to a four-wire circuit, as in a long-distance telephone call, an interface circuit called a hybrid, or terminating, set is used to affect the interface. The hybrid set is used to match impedances and to provide isolation between the two di- rections of signal flow. The hybrid circuit used to convert two-wire circuits to four-wire cir- cuits is similar to the hybrid coil found in standard telephone sets. T-carrier Digital Tx Rx Two-wire Digital T-carrier transmission line Two-wire Two-wire Station B Station A Direction of propagation Unidirectional (c) 3 6 9 # 2 5 8 0 1 4 7 * 3 6 9 # 2 5 8 0 1 4 7 * Two-wire Two-wire T-carrier Digital Radio Transmitter Tx Radio Receiver Rx Direction of Propagation Earth's atmosphere Two-wire Two-wire Station B Station A (d) Unidirectional Transmit antenna Receive antenna 3 6 9 # 2 5 8 0 1 4 7 * 3 6 9 # 2 5 8 0 1 4 7 * Two-wire Two-wire FIGURE 19 (Continued) (c) digital T-carrier system; (d) wireless radio carrier system 432
- 438. Four wire data modem Tx Four wire data modem Rx Rx Tx Four wire passive transmission line Station B Station A (a) Four wire data modem Tx Four wire data modem Rx Rx Tx Four wire active transmission line Station B Station A (b) Amp Amp Four-wire data modem Rx Tx Four-wire data modem Tx Rx Digital T-carrier transceiver Tx Digital T-carrier transceiver Rx Rx Rx Tx Tx Tx Rx Four wire digital T-carrier line Station B Station A (c) Radio transceiver Tx/Rx Radio transceiver Tx/Rx Earth's atmosphere Two-wire Two-wire Station B Station A (d) Bidirectional Transmit/receive antenna Transmit/receive antenna 3 6 9 # 2 5 8 0 1 4 7 * 3 6 9 # 2 5 8 0 1 4 7 * Two-wire Two-wire FIGURE 20 Four-wire configurations: (a) passive cable circuit; (b) active cable circuit; (c) digital T-carrier system; (d) wireless radio carrier system 433
- 439. The Telephone Circuit FIGURE 21 Hybrid (terminating) sets Figure 21 shows the block diagram for a two-wire to four-wire hybrid network. The hy- brid coil compensates for impedance variations in the two-wire portion of the circuit. The am- plifiers and attenuators adjust the signal power to required levels, and the equalizers compen- sate for impairments in the transmission line that affect the frequency response of the transmitted signal, such as line inductance, capacitance, and resistance. Signals traveling west to east (W-E) enter the terminating set from the two-wire line, where they are inductively cou- pled into the west-to-east transmitter section of the four-wire circuit. Signals received from the four-wire side of the hybrid propagate through the receiver in the east-to-west (E-W) section of the four-wire circuit, where they are applied to the center taps of the hybrid coils. If the imped- ances of the two-wire line and the balancing network are properly matched, all currents pro- duced in the upper half of the hybrid by the E-W signals will be equal in magnitude but oppo- site in polarity. Therefore, the voltages induced in the secondaries will be 180° out of phase with each other and, thus, cancel. This prevents any of the signals from being retransmitted to the sender as an echo. If the impedances of the two-wire line and the balancing network are not matched, voltages induced in the secondaries of the hybrid coil will not completely cancel. This im- balance causes a portion of the received signal to be returned to the sender on the W-E por- tion of the four-wire circuit. Balancing networks can never completely match a hybrid to the subscriber loop because of long-term temperature variations and degradation of trans- mission lines. The talker hears the returned portion of the signal as an echo, and if the round-trip delay exceeds approximately 45 ms, the echo can become quite annoying. To eliminate this echo, devices called echo suppressors are inserted at one end of the four-wire circuit. Figure 22 shows a simplified block diagram of an echo suppressor. The speech detec- tor senses the presence and direction of the signal. It then enables the amplifier in the appropri- ate direction and disables the amplifier in the opposite direction, thus preventing the echo 434
- 440. The Telephone Circuit FIGURE 22 Echo suppressor from returning to the speaker. A typical echo suppressor suppresses the returned echo by as much as 60 dB. If the conversation is changing direction rapidly, the people listening may be able to hear the echo suppressors turning on and off (every time an echo suppressor detects speechandisactivated,thefirstinstantofsoundisremovedfromthemessage,givingthespeech a choppy sound). If both parties talk at the same time, neither person is heard by the other. With an echo suppressor in the circuit, transmissions cannot occur in both directions at the same time, thus limiting the circuit to half-duplex operation. Long-distance carriers, such asATT, generally place echo suppressors in four-wire circuits that exceed 1500 elec- trical miles in length (the longer the circuit, the longer the round-trip delay time). Echo sup- pressors are automatically disabled when they receive a tone between 2020 Hz and 2240 Hz, thus allowing full-duplex data transmission over a circuit with an echo suppressor. Full- duplex operation can also be achieved by replacing the echo suppressors with echo cancel- ers. Echo cancelers eliminate the echo by electrically subtracting it from the original signal rather than disabling the amplifier in the return circuit. 7 CROSSTALK Crosstalk can be defined as any disturbance created in a communications channel by sig- nals in other communications channels (i.e., unwanted coupling from one signal path into another). Crosstalk is a potential problem whenever two metallic conductors carrying dif- ferent signals are located in close proximity to each other. Crosstalk can originate in tele- phone offices, at a subscriber’s location, or on the facilities used to interconnect subscriber locations to telephone offices. Crosstalk is a subdivision of the general subject of interfer- ence. The term crosstalk was originally coined to indicate the presence of unwanted speech sounds in a telephone receiver caused by conversations on another telephone circuit. The nature of crosstalk is often described as either intelligible or unintelligible. Intelli- gible(ornearintelligible)crosstalkisparticularlyannoyingandobjectionablebecausethelis- tener senses a real or fancied loss of privacy. Unintelligible crosstalk does not violate privacy, although it can still be annoying. Crosstalk between unlike channels, such as different types of carrier facilities, is usually unintelligible because of frequency inversion, frequency dis- placement, or digital encoding. However, such crosstalk often retains the syllabic pattern of speech and is more annoying than steady-state noise (such as thermal noise) with the same 435
- 441. average power. Intermodulation noise, such as that found in multichannel frequency-division- multiplexed telephone systems, is a form of interchannel crosstalk that is usually unintelligi- ble. Unintelligible crosstalk is generally grouped with other types of noise interferences. The use of the words intelligible and unintelligible can also be applied to non-voice circuits. The methods developed for quantitatively computing and measuring crosstalk be- tween voice circuits are also useful when studying interference between voice circuits and data circuits and between two data circuits. There are three primary types of crosstalk in telephone systems: nonlinear crosstalk, transmittance crosstalk, and coupling crosstalk. 7-1 Nonlinear Crosstalk Nonlinear crosstalk is a direct result of nonlinear amplification (hence the name) in analog communications systems. Nonlinear amplification produces harmonics and cross products (sum and difference frequencies). If the nonlinear frequency components fall into the pass- band of another channel, they are considered crosstalk. Nonlinear crosstalk can be distin- guished from other types of crosstalk because the ratio of the signal power in the disturb- ing channel to the interference power in the disturbed channel is a function of the signal level in the disturbing channel. 7-2 Transmittance Crosstalk Crosstalk can also be caused by inadequate control of the frequency response of a trans- mission system, poor filter design, or poor filter performance. This type of crosstalk is most prevalent when filters do not adequately reject undesired products from other channels. Be- cause this type of interference is caused by inadequate control of the transfer characteris- tics or transmittance of networks, it is called transmittance crosstalk. 7-3 Coupling Crosstalk Electromagnetic coupling between two or more physically isolated transmission media is called coupling crosstalk. The most common coupling is due to the effects of near-field mu- tual induction between cables from physically isolated circuits (i.e., when energy radiates from a wire in one circuit to a wire in a different circuit). To reduce coupling crosstalk due to mutual induction, wires are twisted together (hence the name twisted pair). Twisting the wires causes a canceling effect that helps eliminate crosstalk. Standard telephone cable pairs have 20 twists per foot, whereas data circuits generally require more twists per foot. Direct capacitive coupling between adjacent cables is another means in which signals from one cable can be coupled into another cable. The probability of coupling crosstalk occur- ring increases with cable length, signal power, and frequency. There are two types of coupling crosstalk: near end and far end. Near-end crosstalk (NEXT) is crosstalk that occurs at the transmit end of a circuit and travels in the opposite di- rection as the signal in the disturbing channel. Far-end crosstalk (FEXT) occurs at the far-end receiver and is energy that travels in the same direction as the signal in the disturbing channel. 7-4 Unit of Measurement Crosstalk interference is often expressed in its own special decibel unit of measurement, dBx. Unlike dBm, where the reference is a fixed power level, dBx is referenced to the level on the cable that is being interfered with (whatever the level may be). Mathematically, dBx is dBx 90 (crosstalk loss in decibels) (5) where 90 dB is considered the ideal isolation between adjacent lines. For example, the mag- nitude of the crosstalk on a circuit is 70 dB lower than the power of the signal on the same circuit. The crosstalk is then 90 dB 70 dBx 20 dBx. The Telephone Circuit 436
- 442. QUESTIONS 1. Briefly describe a local subscriber loop. 2. Explain what loading coils and bridge taps are and when they can be detrimental to the per- formance of a telephone circuit. 3. What are the designations used with loading coils? 4. What is meant by the term loop resistance? 5. Briefly describe C-message noise weighting and state its significance. 6. What is the difference between dB and dBm? 7. What is the difference between a TLP and a DLP? 8. What is meant by the following terms: dBmO, rn, dBrn, dBrnc, and dBrncO? 9. What is the difference between psophometric noise weighting and C-message weighting? 10. What are the three categories of transmission parameters? 11. Describe attenuation distortion; envelope delay distortion. 12. What is the reference frequency for attenuation distortion? Envelope delay distortion? 13. What is meant by line conditioning? What types of line conditioning are available? 14. What kind of circuits can have C-type line conditioning; D-type line conditioning? 15. When is D-type conditioning mandatory? 16. What limitations are imposed with D-type conditioning? 17. What is meant by nonlinear distortion? What are two kinds of nonlinear distortion? 18. What considerations are addressed by the interface parameters? 19. What considerations are addressed by facility parameters? 20. Briefly describe the following parameters: 1004-Hz variation, C-message noise, impulse noise, gain hits and dropouts, phase hits, phase jitter, single-frequency interference, frequency shift, phase intercept distortion, and peak-to-average ratio. 21. Describe what is meant by a two-wire circuit; four-wire circuit? 22. Briefly describe the function of a two-wire-to-four-wire hybrid set. 23. What is the purpose of an echo suppressor; echo canceler? 24. Briefly describe crosstalk. 25. What is the difference between intelligible and unintelligible crosstalk? 26. List and describe three types of crosstalk. 27. What is meant by near-end crosstalk; far-end crosstalk? PROBLEMS 1. Describe what the following loading coil designations mean: a. 22B44 b. 19H88 c. 24B44 d. 16B135 2. Frequencies of 250 Hz and 1 kHz are applied to the input of a C-message filter. Would their dif- ference in amplitude be (greater, the same, or less) at the output of the filter? 3. A C-message noise measurement taken at a 22-dBm TLP indicates 72 dBm of noise. A test tone is measured at the same TLP at 25 dBm. Determine the following levels: a. Signal power relative to TLP (dBmO) b. C-message noise relative to reference noise (dBrn) c. C-message noise relative to reference noise adjusted to a 0 TLP (dBrncO) d. Signal-to-noise ratio The Telephone Circuit 437
- 443. 4. A C-message noise measurement taken at a 20-dBm TLP indicates a corrected noise reading of 43 dBrncO. A test tone at data level (0 DLP) is used to determine a signal-to-noise ratio of 30 dB. Determine the following levels: a. Signal power relative to TLP (dBmO) b. C-message noise relative to reference noise (dBrnc) c. Actual test-tone signal power (dBm) d. Actual C-message noise (dBm) 5. A test-tone signal power of 62 dBm is measured at a 61-dBm TLP. The C-message noise is measured at the same TLP at 10 dBrnc. Determine the following levels: a. C-message noise relative to reference noise at a O TLP (dBrncO) b. Actual C-message noise power level (dBm) c. Signal power level relative to TLP (dBmO) d. Signal-to-noise ratio (dB) 6. Sketch the graph for attenuation distortion and envelope delay distortion for a channel with C4 conditioning. 7. An EDD test on a basic telephone channel indicated that a 1600-Hz carrier experienced the min- imum absolute delay of 550 μs. Determine the maximum absolute envelope delay that any fre- quency within the range of 800 Hz to 2600 Hz can experience. 8. The magnitude of the crosstalk on a circuit is 66 dB lower than the power of the signal on the same circuit. Determine the crosstalk in dBx. ANSWERS TO SELECTED PROBLEMS 1. a. 22-gauge wire with 44 mH inductance every 3000 feet b. 19-gauge wire with 88 mH inductance every 6000 feet c. 24-gauge wire with 44 mH inductance every 3000 feet d. 16-gauge wire with 135 mH inductance every 3000 feet 3. a. 3 dBrnO b. 18 dBrnc c. 40 dBrnO d. 47 dB 5. a. 51 dBrncO b. 100 dBm c. 1 dBmO d. 36 dB 7. 2300 μs The Telephone Circuit 438
- 444. The Public Telephone Network CHAPTER OUTLINE 1 Introduction 7 Automated Central Office Switches and Exchanges 2 Telephone Transmission System Environment 8 North American Telephone Numbering Plan Areas 3 The Public Telephone Network 9 Telephone Service 4 Instruments, Local Loops, Trunk Circuits, 10 North American Telephone Switching Hierarchy and Exchanges 11 Common Channel Signaling System No. 7 (SS7) 5 Local Central Office Telephone Exchanges and the Postdivestiture North American 6 Operator-Assisted Local Exchanges Switching Hierarchy OBJECTIVES ■ Define public telephone company ■ Explain the differences between the public and private sectors of the public telephone network ■ Define telephone instruments, local loops, trunk circuits, and exchanges ■ Describe the necessity for central office telephone exchanges ■ Briefly describe the history of the telephone industry ■ Describe operator-assisted local exchanges ■ Describe automated central office switches and exchanges and their advantages over operator-assisted local ex- changes ■ Define circuits, circuit switches, and circuit switching ■ Describe the relationship between local telephone exchanges and exchange areas ■ Define interoffice trunks, tandem trunks, and tandem switches ■ Define toll-connecting trunks, intertoll trunks, and toll offices ■ Describe the North American Telephone Numbering Plan Areas ■ Describe the predivestiture North American Telephone Switching Hierarchy ■ Define the five classes of telephone switching centers ■ Explain switching routes From Chapter 10 of Advanced Electronic Communications Systems, Sixth Edition. Wayne Tomasi. Copyright © 2004 by Pearson Education, Inc. Published by Pearson Prentice Hall. All rights reserved. 439
- 445. ■ Describe the postdivestiture North American Telephone Switching Hierarchy ■ Define Common Channel Signaling System No. 7 (SS7) ■ Describe the basic functions of SS7 ■ Define and describe SS7 signaling points 1 INTRODUCTION The telecommunications industry is the largest industry in the world. There are over 1400 independent telephone companies in the United States, jointly referred to as the public telephone network (PTN). The PTN uses the largest computer network in the world to in- terconnect millions of subscribers in such a way that the myriad of companies function as a single entity. The mere size of the PTN makes it unique and truly a modern-day won- der of the world. Virtually any subscriber to the network can be connected to virtually any other subscriber to the network within a few seconds by simply dialing a telephone num- ber. One characteristic of the PTN that makes it unique from other industries is that every piece of equipment, technique, or procedure, new or old, is capable of working with the rest of the system. In addition, using the PTN does not require any special skills or knowl- edge. 2 TELEPHONE TRANSMISSION SYSTEM ENVIRONMENT In its simplest form, a telephone transmission system is a pair of wires connecting two tele- phones or data modems together. A more practical transmission system is comprised of a complex aggregate of electronic equipment and associated transmission medium, which to- gether provide a multiplicity of channels over which many subscriber’s messages and con- trol signals are propagated. In general, a telephone call between two points is handled by interconnecting a num- ber of different transmission systems in tandem to form an overall transmission path (con- nection) between the two points. The manner in which transmission systems are chosen and interconnected has a strong bearing on the characteristics required of each system because each element in the connection degrades the message to some extent. Consequently, the re- lationship between the performance and the cost of a transmission system cannot be con- sidered only in terms of that system. Instead, a transmission system must be viewed with respect to its relationship to the complete system. To provide a service that permits people or data modems to talk to each other at a dis- tance, the communications system (telephone network) must supply the means and facili- ties for connecting the subscribers at the beginning of a call and disconnecting them at the completion of the call. Therefore, switching, signaling, and transmission functions must be involved in the service. The switching function identifies and connects the subscribers to a suitable transmission path. Signaling functions supply and interpret control and supervisory signals needed to perform the operation. Finally, transmission functions involve the actual transmission of a subscriber’s messages and any necessary control signals. New transmis- sion systems are inhibited by the fact that they must be compatible with an existing multi- trillion-dollar infrastructure. 3 THE PUBLIC TELEPHONE NETWORK The public telephone network (PTN) accommodates two types of subscribers: public and private. Subscribers to the private sector are customers who lease equipment, transmission media (facilities), and services from telephone companies on a permanent basis. The leased The Public Telephone Network 440
- 446. circuits are designed and configured for their use only and are often referred to as private- line circuits or dedicated circuits. For example, large banks do not wish to share their com- munications network with other users, but it is not cost effective for them to construct their own networks. Therefore, banks lease equipment and facilities from public telephone com- panies and essentially operate a private telephone or data network within the PTN. The pub- lic telephone companies are sometimes called service providers, as they lease equipment and provide services to other private companies, organizations, and government agencies. Most metropolitan area networks (MANs) and wide area networks (WANs) utilize private- line data circuits and one or more service provider. Subscribers to the public sector of the PTN share equipment and facilities that are available to all the public subscribers to the network. This equipment is appropriately called common usage equipment, which includes transmission facilities and telephone switches. Anyone with a telephone number is a subscriber to the public sector of the PTN. Since subscribers to the public network are interconnected only temporarily through switches, the network is often appropriately called the public switched telephone network (PSTN) and sometimes simply as the dial-up network. It is possible to interconnect tele- phones and modems with one another over great distances in fractions of a second by means of an elaborate network comprised of central offices, switches, cables (optical and metallic), and wireless radio systems that are connected by routing nodes (a node is a switching point). When someone talks about the public switched telephone network, they are referring to the combination of lines and switches that form a system of electrical routes through the network. In its simplest form, data communications is the transmittal of digital information be- tween two pieces of digital equipment, which includes computers. Several thousand miles may separate the equipment, which necessitates using some form of transmission medium to interconnect them. There is an insufficient number of transmission media capable of car- rying digital information in digital form. Therefore, the most convenient (and least expen- sive) alternative to constructing an all-new all-digital network is to use the existing PTN for the transmission medium. Unfortunately, much of the PTN was designed (and much of it constructed) before the advent of large-scale data communications. The PTN was intended for transferring voice, not digital data. Therefore, to use the PTN for data communications, it is necessary to use a modem to convert the data to a form more suitable for transmission over the wireless carrier systems and conventional transmission media so prevalent in the PTN. There are as many network configurations as there are subscribers in the private sec- tor of the PTN, making it impossible to describe them all. Therefore, the intent of this chap- ter is to describe the public sector of the PTN (i.e., the public switched telephone network). 4 INSTRUMENTS, LOCAL LOOPS, TRUNK CIRCUITS, AND EXCHANGES Telephone network equipment can be broadly divided into four primary classifications: in- struments, local loops, exchanges, and trunk circuits. 4-1 Instruments An instrument is any device used to originate and terminate calls and to transmit and re- ceive signals into and out of the telephone network, such as a 2500-type telephone set, a cordless telephone, or a data modem. The instrument is often referred to as station equip- ment and the location of the instrument as the station. A subscriber is the operator or user of the instrument. If you have a home telephone, you are a subscriber. The Public Telephone Network 441
- 447. 4-2 Local Loops The local loop is simply the dedicated cable facility used to connect an instrument at a sub- scriber’s station to the closest telephone office. In the United States alone, there are sev- eral hundred million miles of cable used for local subscriber loops. Everyone who sub- scribes to the PTN is connected to the closest telephone office through a local loop. Local loops connected to the public switched telephone network are two-wire metallic cable pairs. However, local loops used with private-line data circuits are generally four-wire configurations. 4-3 Trunk Circuits A trunk circuit is similar to a local loop except trunk circuits are used to interconnect two telephone offices. The primary difference between a local loop and a trunk is that a local loop is permanently associated with a particular station, whereas a trunk is a common- usage connection.A trunk circuit can be as simple as a pair of copper wires twisted together or as sophisticated as an optical fiber cable. A trunk circuit could also be a wireless com- munications channel. Although all trunk circuits perform the same basic function, there are different names given to them, depending on what types of telephone offices they inter- connect and for what reason. Trunk circuits can be two wire or four wire, depending on what type of facility is used. Trunks are described in more detail in a later section of this chapter. 4-4 Exchanges An exchange is a central location where subscribers are interconnected, either temporar- ily or on a permanent basis. Telephone company switching machines are located in ex- changes. Switching machines are programmable matrices that provide temporary signal paths between two subscribers. Telephone sets and data modems are connected through local loops to switching machines located in exchanges. Exchanges connected directly to local loops are often called local exchanges or sometimes dial switches or local dial switches. The first telephone exchange was installed in 1878, only two years after the in- vention of the telephone. A central exchange is also called a central telephone exchange, central office (CO), central wire center, central exchange, central office exchange, or sim- ply central. The purpose of a telephone exchange is to provide a path for a call to be completed between two parties. To process a call, a switch must provide three primary functions: Identify the subscribers Set up or establish a communications path Supervise the calling processes 5 LOCAL CENTRAL OFFICE TELEPHONE EXCHANGES The first telephone sets were self-contained, as they were equipped with their own battery, microphone, speaker, bell, and ringing circuit. Telephone sets were originally connected di- rectly to each other with heavy-gauge iron wire strung between poles, requiring a dedicated cable pair and telephone set for each subscriber you wished to be connected to. Figure 1a shows two telephones interconnected with a single cable pair. This is simple enough; how- ever, if more than a few subscribers wished to be directly connected together, it became cumbersome, expensive, and very impractical. For example, to interconnect one subscriber to five other subscribers, five telephone sets and five cable pairs are needed, as shown in Figure 1b. To completely interconnect four subscribers, it would require six cable pairs, and each subscriber would need three telephone sets. This is shown in Figure 1c. The Public Telephone Network 442
- 448. Single cable pair Telephone 1 Telephone 2 (a) Telephone 1 Telephone 2 Telephone 3 Telephone 4 Telephone 5 (b) One subscriber's location Cable pair 1 Cable pair 2 Cable pair 3 Cable pair 4 Cable pair 5 The number of lines required to interconnect any number of stations is determined by the following equation: (1) where n number of stations (parties) N number of interconnecting lines The number of dedicated lines necessary to interconnect 100 parties is N 1001100 12 2 4950 N n1n 12 2 The Public Telephone Network FIGURE 1 Dedicated telephone interconnections: (a) Interconnecting two subscribers; (b) Interconnecting one subscriber to five other telephone sets; (Continued) 443
- 449. Subscriber 1 Subscriber 2 Subscriber 4 Six Cable pairs Subscriber 3 (c) In addition, each station would require either 100 separate telephones or the capability of switching one telephone to any of 99 lines. These limitations rapidly led to the development of the central telephone exchange. A telephone exchange allows any telephone connected to it to be interconnected to any of the other telephones connected to the exchange without requiring separate cable pairs and telephones for each connection. Generally, a community is served by only one telephone company. The community is divided into zones, and each zone is served by a different cen- tral telephone exchange. The number of stations served and the density determine the num- ber of zones established in a given community. If a subscriber in one zone wishes to call a station in another zone, a minimum of two local exchanges is required. 6 OPERATOR-ASSISTED LOCAL EXCHANGES The first commercial telephone switchboard began operation in New Haven, Connecticut, on January 28, 1878, marking the birth of the public switched telephone network. The switchboard served 21 telephones attached to only eight lines (obviously, some were party lines). On February 17 of the same year, Western Union opened the first large-city exchange in San Francisco, California, and on February 21, the New Haven District Telephone Com- pany published the world’s first telephone directory comprised of a single page listing only 50 names. The directory was immediately followed by a more comprehensive listing by the Boston Telephone Dispatch Company. The first local telephone exchanges were switchboards (sometimes called patch pan- els or patch boards) where manual interconnects were accomplished using patchcords and The Public Telephone Network FIGURE 1 (Continued) (c) Interconnecting four subscribers 444
- 450. Local loop Local loop Lamps Jacks Local loop Telephone set Telephone set Telephone set Relay Relay Relay jacks. All subscriber stations were connected through local loops to jacks on the switch- board. Whenever someone wished to initiate a call, they sent a ringing signal to the switch- board by manually turning a crank on their telephone. The ringing signal operated a relay at the switchboard, which in turn illuminated a supervisory lamp located above the jack for that line, as shown in Figure 2. Manual switchboards remained in operation until 1978, when the Bell System replaced their last cord switchboard on Santa Catalina Island off the coast of California near Los Angeles. In the early days of telephone exchanges, each telephone line could have 10 or more subscribers (residents) connected to the central office exchange using the same local loop. This is called a party line, although only one subscriber could use their telephone at a time. Party lines are less expensive than private lines, but they are also less convenient. A private telephone line is more expensive because only telephones from one residence or business are connected to a local loop. Connecting 100 private telephone lines to a single exchange required 100 local loops and a switchboard equipped with 100 relays, jacks, and lamps. When someone wished to initiate a telephone call, they rang the switchboard. An operator answered the call by saying, “Central.” The calling party told the operator whom they wished to be connected to. The operator would then ring the destination, and when someone an- swered the telephone, the operator would remove her plug from the jack and connect the calling and called parties together with a special patchcord equipped with plugs on both ends. This type of system was called a ringdown system. If only a few subscribers were connected to a switchboard, the operator had little trouble keeping track of which jacks were for which subscriber (usually by name). However, as the popularity of the tele- phone grew, it soon became necessary to assign each subscriber line a unique telephone number. A switchboard using four digits could accommodate 10,000 telephone numbers (0000 to 9999). Figure 3a shows a central office patch panel connected to four idle subscriber lines. Note that none of the telephone lines is connected to any of the other telephone lines. Figure 3b shows how subscriber 1 can be connected to subscriber 2 using a tem- porary connection provided by placing a patchcord between the jack for line 1 and the jack for line 2. Any subscriber can be connected to any other subscriber using patch- cords. The Public Telephone Network FIGURE 2 Patch panel configuration 445
- 451. Telephone subscriber 3 Patch panel with no crossconnects between subscribers 3 6 9 # 2 5 8 0 1 4 7 * Telephone subscriber 1 3 6 9 # 2 5 8 0 1 4 7 * Telephone subscriber 4 3 6 9 # 2 5 8 0 1 4 7 * Telephone subscriber 2 3 6 9 # 2 5 8 0 1 4 7 * (a) Telephone subscriber 3 Patch panel with connection between subscriber 1 and subscriber 2 3 6 9 # 2 5 8 0 1 4 7 * Telephone subscriber 1 3 6 9 # 2 5 8 0 1 4 7 * Telephone subscriber 4 3 6 9 # 2 5 8 0 1 4 7 * Telephone subscriber 2 3 6 9 # 2 5 8 0 1 4 7 * (b) FIGURE 3 Central office exchange: (a) without interconnects; (b) with an interconnect 7 AUTOMATED CENTRAL OFFICE SWITCHES AND EXCHANGES As the number of telephones in the United States grew, it quickly became obvious that operator-assisted calls and manual patch panels could not meet the high demand for ser- vice. Thus, automated switching machines and exchange systems were developed. Anautomatedswitchingsystemisasystemofsensors,switches,andotherelectricaland electronic devices that allows subscribers to give instructions directly to the switch without havingtogothroughanoperator.Inaddition,automatedswitchesperformedinterconnections between subscribers without the assistance of a human and without using patchcords. In 1890 an undertaker in Kansas City, Kansas, named Alman Brown Strowger was concerned that telephone company operators were diverting his business to his competitors. Consequently, he invented the first automated switching system using electromechanical relays. It is said that Strowger worked out his original design using a cardboard box, straight pins, and a pencil. The Public Telephone Network 446
- 452. With the advent of the Strowger switch, mechanical dialing mechanisms were added to the basic telephone set. The mechanical dialer allowed subscribers to manually dial the telephone number of the party they wished to call. After a digit was entered (dialed), a re- lay in the switching machine connected the caller to another relay. The relays were called stepping relays because the system stepped through a series of relays as the digits were en- tered. The stepping process continued until all the digits of the telephone number were en- tered. This type of switching machine was called a step-by-step (SXS) switch, stepper, or, perhaps more commonly, a Strowger switch. A step-by-step switch is an example of a pro- gressive switching machine, meaning that the connection between the calling and called parties was accomplished through a series of steps. Between the early 1900s and the mid-1960s, the Strowger switch gradually replaced manual switchboards. The Bell System began using steppers in 1919 and continued using them until the early 1960s. In 1938, the Bell System began replacing the steppers with an- other electromechanical switching machine called the crossbar (XBAR) switch. The first No. 1 crossbar was cut into service at the Troy Avenue central office in Brooklyn, New York, on February 14, 1938. The crossbar switch used sets of contact points (called cross- points) mounted on horizontal and vertical bars. Electromagnets were used to cause a ver- tical bar to cross a horizontal bar and make contact at a coordinate determined by the called number. The most versatile and popular crossbar switch was the #5XB. Although crossbar switches were an improvement over step-by-step switches, they were short lived, and most of them have been replaced with electronic switching systems (ESS). In 1965, ATT introduced the No. 1 ESS, which was the first computer-controlled central office switching system used on the PSTN. ESS switches differed from their pre- decessors in that they incorporate stored program control (SPC), which uses software to control practically all the switching functions. SPC increases the flexibility of the switch, dramatically increases its reliability, and allows for automatic monitoring of maintenance capabilities from a remote location. Virtually all the switching machines in use today are electronic stored program control switching machines. SPC systems require little mainte- nance and require considerably less space than their electromechanical predecessors. SPC systems make it possible for telephone companies to offer the myriad of services available today, such as three-way calling, call waiting, caller identification, call forwarding, call within, speed dialing, return call, automatic redial, and call tracing. Electronic switching systems evolved from the No. 1 ESS to the No. 5 ESS, which is the most advanced digital switching machine developed by the Bell System. Automated central office switches paved the way for totally automated central office exchanges, which allow a caller located virtually anywhere in the world to direct dial vir- tually anyone else in the world. Automated central office exchanges interpret telephone numbers as an address on the PSTN. The network automatically locates the called number, tests its availability, and then completes the call. 7-1 Circuits, Circuit Switches, and Circuit Switching A circuit is simply the path over which voice, data, or video signals propagate. In telecom- munications terminology, a circuit is the path between a source and a destination (i.e., be- tween a calling and a called party). Circuits are sometimes called lines (as in telephone lines). A circuit switch is a programmable matrix that allows circuits to be connected to one another. Telephone company circuit switches interconnect input loop or trunk circuits to output loop or trunk circuits. The switches are capable of interconnecting any circuit connected to it to any other circuit connected to it. For this reason, the switching process is called circuit switching and, therefore, the public telephone network is considered a circuit-switched net- work. Circuit switches are transparent. That is, they interconnect circuits without altering the information on them. Once a circuit switching operation has been performed, a transparent switch simply provides continuity between two circuits. The Public Telephone Network 447
- 453. Local loops Local loops Local loops Local Exchange 874-3333 874-2222 3 6 9 # 2 5 8 0 1 4 7 * 874-6666 3 6 9 # 2 5 8 0 1 4 7 * 874-7777 3 6 9 # 2 5 8 0 1 4 7 * 3 6 9 # 2 5 8 0 1 4 7 * 874-5555 874-4444 3 6 9 # 2 5 8 0 1 4 7 * 3 6 9 # 2 5 8 0 1 4 7 * (a) Local loops Local loops Local loops Local Exchange 874-6666 3 6 9 # 2 5 8 0 1 4 7 * 874-7777 3 6 9 # 2 5 8 0 1 4 7 * 874-5555 3 6 9 # 2 5 8 0 1 4 7 * 874-3333 3 6 9 # 2 5 8 0 1 4 7 * 874-4444 3 6 9 # 2 5 8 0 1 4 7 * 874-2222 3 6 9 # 2 5 8 0 1 4 7 * (b) FIGURE 4 Local exchange: (a) no interconnections; (b) 874-3333 connected to 874-4444 7-2 Local Telephone Exchanges and Exchange Areas Telephone exchanges are strategically placed around a city to minimize the distance be- tween a subscriber’s location and the exchange and also to optimize the number of stations connected to any one exchange. The size of the service area covered by an exchange de- pends on subscriber density and subscriber calling patterns. Today, there are over 20,000 local exchanges in the United States. Exchanges connected directly to local loops are appropriately called local exchanges. Because local exchanges are centrally located within the area they serve, they are often called central offices (CO). Local exchanges can directly interconnect any two subscribers whose local loops are connected to the same local exchange. Figure 4a shows a local ex- change with six telephones connected to it. Note that all six telephone numbers begin with 87. One subscriber of the local exchange can call another subscriber by simply dialing their seven-digit telephone number. The switching machine performs all tests and switching op- erations necessary to complete the call.A telephone call completed within a single local ex- change is called an intraoffice call (sometimes called an intraswitch call). Figure 4b shows how two stations serviced by the same exchange (874-3333 to 874-4444) are intercon- nected through a common local switch. In the days of manual patch panels, to differentiate telephone numbers in one local ex- change from telephone numbers in another local exchange and to make it easier for people to The Public Telephone Network 448
- 454. remember telephone numbers, each exchange was given a name, such as Bronx, Redwood, Swift,Downtown,Main,andsoon.Thefirsttwodigitsofatelephonenumberwerederivedfrom the first two letters of the exchange name. To accommodate the names with dial telephones, the digits 2 through 9 were each assigned three letters. Originally, only 24 of the 26 letters were as- signed (Q and Z were omitted); however, modern telephones assign all 26 letters to oblige per- sonalizing telephone numbers (the digits 7 and 9 are now assigned four letters each). As an ex- ample, telephone numbers in the Bronx exchange begin with 27 (B on a telephone dial equates to the digit 2, and R on a telephone dial equates to the digit 7). Using this system, a seven-digit telephone number can accommodate 100,000 telephone numbers. For example, the Bronx ex- change was assigned telephone numbers between 270-0000 and 279-9999 inclusive. The same 100,000 numbers could also be assigned to the Redwood exchange (730-0000 to 739-9999). 7-3 Interoffice Trunks, Tandem Trunks, and Tandem Switches Interoffice calls are calls placed between two stations that are connected to different local exchanges. Interoffice calls are sometimes called interswitch calls. Interoffice calls were originally accomplished by placing special plugs on the switchboards that were connected to cable pairs going to local exchange offices in other locations around the city or in nearby towns. Today telephone-switching machines in local exchanges are interconnected to other local exchange offices on special facilities called trunks or, more specifically, interoffice trunks. A subscriber in one local exchange can call a subscriber connected to another local exchange over an interoffice trunk circuit in much the same manner that they would call a subscriber connected to the same exchange. When a subscriber on one local exchange di- als the telephone number of a subscriber on another local exchange, the two local exchanges are interconnected with an interoffice trunk for the duration of the call. After either party terminates the call, the interoffice trunk is disconnected from the two local loops and made available for another interoffice call. Figure 5 shows three exchange offices with The Public Telephone Network Local loops Interoffice trunk BRonx Exchange 27X-XXXX Interoffice trunks Local loops 274-1234 274-4321 SWift Exchange 79X-XXXX 795-6789 795-9876 Local loops UPtown Exchange 87X-XXXX 873-4567 873-7654 FIGURE 5 Interoffice exchange system 449
- 455. two subscribers connected to each. The telephone numbers for subscribers connected to the Bronx, Swift, and Uptown exchanges begin with the digits 27, 79, and 87, respectively. Figure 6 shows how two subscribers connected to different local exchanges can be inter- connected using an interoffice trunk. In larger metropolitan areas, it is virtually impossible to provide interoffice trunk cir- cuits between all the local exchange offices. To interconnect local offices that do not have interoffice trunks directly between them, tandem offices are used. A tandem office is an ex- change without any local loops connected to it (tandem meaning “in conjunction with” or “associated with”). The only facilities connected to the switching machine in a tandem of- fice are trunks. Therefore, tandem switches interconnect local offices only.A tandem switch is called a switcher’s switch, and trunk circuits that terminate in tandem switches are ap- propriately called tandem trunks or sometimes intermediate trunks. Figure 7 shows two exchange areas that can be interconnected either with a tandem switch or through an interoffice trunk circuit. Note that tandem trunks are used to connect the Bronx and Uptown exchanges to the tandem switch. There is no name given to the tandem switch because there are no subscribers connected directly to it (i.e., no one receives dial tone from the tandem switch). Figure 8 shows how a subscriber in the Uptown exchange area is connected to a subscriber in the Bronx exchange area through a tandem switch.As the figure shows, tandem offices do not eliminate interoffice trunks. Very often, local offices have the capabilities to be interconnected with direct interoffice trunks as well as through a tandem of- fice.When a telephone call is made from one local office to another, an interoffice trunk is se- lected if one is available. If not, a route through a tandem office is the second choice. The Public Telephone Network Local loops Interoffice trunk BRonx Exchange 27X-XXXX Interoffice trunks Local loops 274-1234 274-4321 SWift Exchange 79X-XXXX 795-6789 Local loops UPtown Exchange 87X-XXXX 873-4567 873-7654 795-9876 FIGURE 6 Interoffice call between subscribers serviced by two different exchanges 450
- 456. Local loops Tandem trunk BRonx Exchange 27X-XXXX Interoffice trunk Tandem trunk 275-0001 275-0002 Tandem Switch Local loops UPtown Exchange 87X-XXXX 874-0002 874-0001 Local loops Tandem trunk BRonx Exchange 27X-XXXX Interoffice trunk Tandem trunk 275-0003 Tandem Switch Local loops UPtown Exchange 87X-XXXX 874-2209 874-2222 275-9956 FIGURE 8 Interoffice call between two local exchanges through a tandem switch FIGURE 7 Interoffice switching between two local exchanges using tandem trunks and a tandem switch 451
- 457. Local exchange office Local exchange office To other local offices To other toll offices To other toll offices Local exchange office Subscribers Local loops Local exchange office Subscribers Local loops Subscribers Interoffice trunks Intertoll trunks Toll connecting trunks Toll connecting trunk Toll connecting trunk Local loops Subscribers Local loops Toll exchange office Toll exchange office The Public Telephone Network 7-4 Toll-Connecting Trunks, Intertoll Trunks, and Toll Offices Interstate long-distance telephone calls require a special telephone office called a toll office. There are approximately 1200 toll offices in the United States. When a subscriber initiates a long-distance call, the local exchange connects the caller to a toll office through a facility called a toll-connecting trunk (sometimes called an interoffice toll trunk). Toll offices are connected to other toll offices with intertoll trunks. Figure 9 shows how local exchanges are connected to toll offices and how toll offices are connected to other toll offices. Figure 10 shows the network relationship between local exchange offices, tandem offices, toll offices, and their respective trunk circuits. 8 NORTH AMERICAN TELEPHONE NUMBERING PLAN AREAS The North American Telephone Numbering Plan (NANP) was established to provide a tele- phone numbering system for the United States, Mexico, and Canada that would allow any subscriber in North America to direct dial virtually any other subscriber without the assis- tance of an operator. The network is often referred to as the DDD (direct distance dialing) network. Prior to the establishment