The fundamentals of sonet

Technical Note

The Fundamentals of SONET
Overview

When fiber optical cables were initially deployed as a medium for high-speed digital transport, the lack
of standards led to widespread deployment of proprietary optical interfaces. This meant that fiber optic
transmission equipment from one manufacturer could not interface with equipment from any of the
other manufacturers. Service providers were required to select a single vendor for deployment
throughout the network and then were locked in to the network control and monitoring capabilities of
that manufacturer. Although this technology satisfied the bandwidth needs of the network for several
years, it was evident that this arrangement could not support the future needs of the industry because
of the limited interconnection capabilities.
In 1985, Bellcore proposed the idea of an optical carrier-to-carrier interface that would allow the
interconnection of different manufacturers’ optical equipment. This was based on a hierarchy of digital
rates, all formed by the interleaving of a basic rate signal. The idea of a Synchronous Optical NETwork
(SONET) attracted the interest of carriers, Regional Bell Operating Companies (RBOCs), and
manufacturers alike and quickly gained momentum. Interest in SONET by CCITT (now International
Telecommunication Union – ITU-T) expanded its scope from a domestic to an international standard,
and by 1988 the ANSI committee had successfully integrated changes requested by the ITU-T, and were
well on their way toward the issuance of the new standard. Today, the SONET standard is contained in
the ANSI specification T1.105 Digital Hierarchy – Optical Interface Rates & Formats Specifications
(SONET), and technical recommendations are found in Bellcore GR-253-CORE Synchronous Optical
Network (SONET) Transport Systems: Common Generic Criteria.
The SONET specifications define optical carrier (OC) interfaces and their electrical equivalents to allow
transmission of lower-rate signals at a common synchronous rate. One of the benefits of the SONET
signal, as with any standard, is that it allows multiple vendors to provide compatible transmission
equipment in the same span. SONET also allows for dynamic drop and insert capabilities on the
payload without the delay and additional hardware associated with demultiplexing and remultiplexing
the higher rate signal. Since the overhead is relatively independent of the payload, SONET is able to
integrate new services, such as Asynchronous Transfer Mode (ATM) and Fiber Distributed Data
Interface (FDDI), in addition to existing DS3 and DS1 services. Another major advantage of SONET is
that the operations, administration, maintenance, and provisioning (OAM&P) capabilities are built
directly into the signal overhead to allow maintenance of the network from one central location.

SONET Multiplexing

SONET multiplexing combines low-speed digital signals such as DS1, DS1C, E1, DS2, and DS3 with
required overhead to form a building block called Synchronous Transport Signal Level One (STS-1).
Figure 1 shows the STS-1 frame, which is organized as 9 rows by 90 columns of bytes. It is transmitted
row first, with the most significant bit (MSB) of each byte transmitted first.

WEBSITE : www.jdsu.com

Technical Note: The Fundamentals of SONET 2

90 Bytes

B B B 87B

9
Rows

125 µs
Transport STS-1 Envelope Capacity
Overhead

B denotes an 8-bit byte

Figure 1 STS-1 frame

A generic formula calculates the bit rate of a framed digital signal:
bit rate = frame rate x frame capacity
In order for SONET to easily integrate existing digital services into its hierarchy, it was defined to
operate at the basic rate of 8 kHz or 125 microseconds per frame, so the frame rate is 8,000 frames per
second.
The frame capacity of a signal is the number of bits contained within a single frame. Figure 1 shows:
frame capacity = 90 bytes/row x 9 rows/frame x 8 bits/byte = 6,480 bits/frame
Now the bit rate of the STS-1 signal is calculated as follows:
bit rate = 8,000 frames/second x 6,480 bits/frame = 51.840 Mbps
Higher-rate signals are formed by combining multiples of the STS-1 block by interleaving a byte from
each STS-1 to form an STS-3, as shown in Figure 2. The basic frame rate remains 8,000 frames per
second, but the capacity is tripled to result in a bit rate of 155.52 Mbps. The STS-3 may then be
converted to an optical signal (OC-3) for transport, or further multiplexed with three additional STS-
3s to form an STS-12 signal, and so on. Table 1 defines common SONET optical rates, their equivalent
electrical rates, and the maximum number of DS0 voice channels which can be carried at that rate.

STS-1

C B A C B A
STS-1 3:1
STS-3

STS-1

Figure 2 Multiplexing STS-1s


Frame Format Optical Bit Rate Maximum DS0s
SONET Rates
STS-1 OC-1 51.84 Mbps 672*
STS-3 OC-3** 155.52 Mbps 2,016
STS-12 OC-12** 622.08 Mbps 8,064
STS-24 OC-24 1.244 Gbps 16,128
STS-48 OC-48** 2.488 Gbps 32,256
STS-192 OC-192 9.953 Gbps 129,024
*Same number of DS0s as a DS3 signal
**Most popular transport interfaces today

SONET Frame

Figure 3 shows the STS-1 frame divided into two parts to physically segregate the layers, where each
square represents an 8-bit byte. The first three columns comprise the transport overhead (TOH), while
the remainder is called the synchronous payload envelope (SPE). The TOH dedicates three rows for the
section overhead (SOH) and six rows for the line overhead (LOH). The SPE contains one column for
the path overhead (POH), leaving the remaining 86 columns for payload data (49.536 Mbps). Appendix
A is included as a reference to describe each of the bytes in Figure 3.

Transport Overhead Synchronous Payload Envelope

Framing Framing Trace Trace
A1 A2 J0 J1

Section BIP-8 Orderwire User BIP-8
Overhead B1 E1 F1 B3

Data Com Data Com Data Com Signal Label
D1 D2 D3 C2

Pointer Pointer Pointer Action Path Status
H1 H2 H3 G1
9
BIP-8 APS APS Rows User Channel
B2 K1 K2 F2

Data Com Data Com Data Com Multiframe
Line D4 D5 D6 H4
Overhead

Data Com Data Com Data Com Growth
D7 D8 D9 Z3

Data Com Data Com Data Com Growth
D10 D11 D12 Z4

Synch FEBE Orderwire Tandem Connect
S1 M0 E2 Z5

3 Bytes 87 Bytes

Figure 3 SONET overhead structure


SONET Signal Hierarchy

The SONET signal is layered to divide responsibility for transporting the payload through the network.
Each network element (NE) is responsible for interpreting and generating its overhead layer, and for
communicating control and status information to the same layer in other equipment – in short,
“terminating” its overhead layer. As the payload travels through the SONET network, each layer is
terminated by one of a general class of NEs, termed section terminating equipment (STE), line
terminating equipment (LTE), or path terminating equipment (PTE). Figure 4 illustrates a sample
network with the layered functions identified. The POH is generated at the point where the lower-rate
signal enters the SONET network by PTE such as a terminal multiplexer (TM). The POH is removed
when the payload exits the network. Since the POH is first-on last-off, alarm and error information
contained within this layer represents end-to-end status.
Path Layer
Line Layer Line Layer
West 1 2 4 East

OC-N OC-N OC-N 3 OC-N
DSn DSn
TM REGEN ADM REGEN TM

CPE CPE

Section Layer Section Layer Section Layer Section Layer

TM = Terminal Multiplexer
CPE = Customer Premises Equipment
REGEN = Optical Regenerator
ADM = Add/Drop Multiplexer

Figure 4 Typical layered communication network

The next layer of overhead termination is the LOH and is performed by the LTE such as a SONET
add/drop multiplexer (ADM). The LOH is where most of the communication and synchronization
between NEs occurs, and represents error information between major nodes in the network. Finally,
SOH is terminated by STE, such as optical regenerators, and contains error information between every
node in the network. In many cases, LTE, PTE, and STE functions are combined within the identical
piece of equipment. Since each layer is terminated and regenerated at the appropriate nodes, the
performance monitoring data at each NE will help to sectionalize a problem.
For example, if traffic is traveling west to east in Figure 4, and section errors are detected at Site 4, a
problem will be somewhere between Site 3 and Site 4. The observed problem cannot be west of Site 3,
since all section results are recalculated at every point in the network. If line errors are found at Site 4,
a problem exists between Site 2 and Site 4, since line results are recalculated only at major nodes in the
network, such as the ADM at Site 2. Finally, if path errors are detected at Site 4, then a problem exists
anywhere between Site 1 and Site 4.
The ADM at Site 2 adds a twist to the path errors example, due to its flexible functionality as shown in
Figure 5. An ADM functions as a PTE when the signals being dropped and added are tributaries of the
SONET signal. If the ADM has been equipped to add and drop DS3 or DS1 signals, the ADM functions
as a PTE for those signals. If it is equipped to add and drop STS-1 or OC-n signals, the ADM functions
only as an LTE for those signals. This fact must be considered in the scenario in Figure 4. The path
statement must be modified to add the condition that if path errors are located at Site 4, and the origin
of a DSn tributary within the STS-1 is at Site 1, then a problem exists between Site 1 and Site 4.
Otherwise, if any Dsn tributary within the STS-1 originates from Site 2, then a problem exists between
Site 2 and Site 4. So, in troubleshooting a signal, it is important to know where the path originates.


Terminal Terminal
Multiplexer Multiplexer

STS-1, OC-N
OC-N STS-1 STS-1 OC-N
DSn DSn

Signal is section, line,
and path terminated Signals are section and line
(ADM is PTE) terminated (ADM is TE)

Figure 5 Function of a SONET add/drop multiplexer

Since the origin of the signal is an important factor used to isolate trouble spots, the SONET signal itself
provides a method to tag every STS-1 with information about its location. The J1 Path Trace Byte in
Appendix A fills this role. This byte repetitively carries a fixed-length, 64-byte, ASCII string that can be
programmed at system turn-up to carry textual information about the originating node, office, or
customer. Because this information is never terminated by LTE or STE, it can only be assigned at the
originating point of the signal.

SONET Performance Monitoring

Each layer in the SONET signal provides alarm and error monitoring capabilities between various
terminating points in the network. Similar to DS3 and DS1 signals, parity is calculated and stored in
the transmitted signal. The parity is recalculated by the receiver and verified against the stored value to
determine if an error occurred during transmission. Every layer in the SONET signal has its own Bit
Interleaved Parity (BIP) calculation. The sidebar on the next page shows how BIP checks are performed
in SONET.
When an error is detected in a C-bit DS3 signal, a remote error indication (REI), formerly referred to
as far-end block error (FEBE), is returned to the sender. SONET uses the same algorithm, using a
layered approach. If an LTE receives some number of line BIP errors, it transmits the same number of
REI-L errors back to the originator. PTE use the same approach in the path layer of overhead. The
SONET signal also contains AIS and Yellow alarms, like DS3 and DS1, except a SONET Yellow alarm is
called a remote defect indication (RDI), and is also layered like all of the other SONET results. The term
RDI replaces the former names FERF (far-end receive failure) and RAI (remote alarm indication) from
previous versions of the SONET specification.


Bit Interleaved Parity (BIP)
BIP calculations are performed over each layer of the SONET overhead, such that each bit in the
BIP byte will indicate the parity of all respective bits in the previous frame. For example, if the
number of bits equaling one in the first bit position of every byte is odd, then the first bit
position of the BIP byte will be one. If the number of ones in the first position is even, then the
first bit position of the BIP byte will be zero. This is repeated for all eight bits of each byte to
determine the value of the BIP byte.
Bytes in Transmitted Signal = 0110 0100
1010 1110
BIP Calculation = 1100 1010
Each layer calculates the BIP for all information in its domain. For example, since the entire
SONET signal is formed when the STE sees it, the section BIP is calculated over the entire signal,
including all SOH, LOH, and POH of the previous STS-n frame. The result is then placed in the
B1 byte for STS-1 Number 1 of an STS-n. Line BIPs are calculated over the previous STS-1 frame,
minus the SOH, and placed in the B2 byte for every STS-1 of an STS-n. Path BIPs are calculated
over the previous frame, minus SOH and LOH, and are found in the B3 byte of every STS-1 of
an STS-n.

SONET Timing Compensation
The SONET signal was designed to be timing tolerant to support asynchronously timed, lower-rate
signals and slight timing differences between synchronously timed NEs. There are two mechanisms
which allow for robust timing compensation: variable bit stuffing of the lower-rate signal, and a
technique called “pointer adjustments” between synchronous elements in the SONET network.

Pointer Adjustments
Pointer adjustments allow the SPE to “float” with respect to the SONET frame. This means that a single
SPE payload frame typically crosses the STS-1 frame boundary, as shown in Figure 6. The “pointer” is
contained in the H1 and H2 bytes of the LOH and is a count of the number of bytes the J1 byte is away
from the H3 byte, not including the TOH bytes. A valid pointer can range from 0 to 782.
When timing differences exist, dummy bytes can be inserted into the SPE without affecting the data.
Since the pointer is adjusted to indicate where the real POH starts, the receiving end can effectively
recover the payload (i.e., ignore the dummy bytes). When “stuffed bytes” are used, they are always in
the same location, regardless of where the POH starts. H3 is called a “negative stuff byte” and is used to
carry real payload data for one frame during a pointer decrement. The byte following H3 in the SPE is
called a “positive stuff byte” and is used to carry a dummy byte of information for one frame during a
pointer increment.


3 Bytes 87 Bytes

H1 H2 H3

STS-1 SPE
J1

9
Rows

Frame N

STS-1
POH

125 µs

H1 H2 H3 Frame N + 1

9
Rows

250 µs

Figure 6 Pointer bytes designating the start of the SPE path overhead

If there is no timing difference between two nodes, the incoming STS-1 payload bit rate is identical to
the transmit timing source that drives the outgoing STS-1 frame rate, so no pointer adjustments are
needed. Figure 7 shows a SONET node which has an incoming frequency f1 and an outgoing frequency
f2. If f1 is less than f2, there is a constant lack of payload data to place into the outgoing SONET signal.
To compensate, a dummy byte is placed into the positive stuff byte and all the data is moved right by
one byte, so the pointer is incremented by one (Figure 8). On the other hand, if f1 is greater than f2 as
shown in Figure 9, then an extra SPE payload byte is stored into the negative stuff byte, H3, in the LOH
for one frame, while all the payload data is moved left by one byte and the pointer is decremented by
one (Figure 10).

f2 SONET f1
Node (LTE)
f1< f2

Figure 7 Node with slower incoming data rate


STS-1 Frame
0 µs

Pointer Value (P)
H1 H2 H3
Start of STS-1 SPE
Frame N

125 µs

P
H1 H2 H3 Positive Stuff Byte
(Dummy Byte)

Frame N + 1

250 µs

PNEW = P + 1
H1 H2 H3

Frame N + 2

375 µs

Figure 8 Incrementing the pointer value

f2 SONET f1
Node (LTE)
f1> f2

Figure 9 Node with faster incoming data rate


STS-1 Frame
0 µs

Pointer Value (P)
H1 H2 H3
Start of STS-1 SPE
Frame N

125 µs

P
H1 H2 Negative Stuff Byte (Data)

Frame N + 1

250 µs

PNEW = P-1
H1 H2 H3

Frame N + 2

375 µs

Figure 10 Decrementing the pointer value

An LTE is the only equipment which can perform path pointer adjustments, since the pointer value is
contained in the LOH. Also, path pointer adjustments are not performed by PTE, where the payload
data enters the SONET network, even though there are potential timing differences at these locations
as well. The timing differences at PTE are due to asynchronously-timed tributary signals and are
corrected using traditional bit stuffing techniques.
Figure 11 shows a simplified version of how a SONET signal is assembled and disassembled to
summarize the layered responsibilities in SONET. All of the steps occur in a single 125 microsecond
period over a single SONET frame.


Transmit Functions Receive Functions
Path Layer
1 Append the path overhead to 8 Calculate the path BIP over
the tributary signal, using the the SPE and store for use in
previously calculated path BIP next frame. Separate the path
in the B3 byte. overhead from the payload so
it can be dropped as needed.

2 Calculate the new path BIP 7 Examine B3 byte and
value over the SPE and store compare it to last saved
for use in the next frame. REI-P version of path BIPs. Report
path BIP errors if
discrepancies are found.

Line Layer
3 Assign a path pointer to the 6 Calculate a BIP over the line
newly formed SPE in the line overhead and the SPE. Save it
overhead (LOH). for use in the next frame.

4 Fill in the rest of the LOH using 5 Examine B2 byte and
the previously calculated line compare it to last saved
BIP in B2. REI-L version of the line BIPs.
Report line BIP errors if any
found.

5 Calculate the new line BIP 4 Examine the pointer value to
value over the LOH and SPE find the start of the SPE.
and store for use in the next Handle pointer adjustments.
frame.

Section Layer
6 Insert all section overhead 3 Descramble all of the data by
(SOH) except A1, A2, C1. Use Exclusive-Oring with the 2^7-1
the previous value of section PRBS. Compare B1 byte to
BIP for the B1 location. previously saved value and
report section BIP errors if
discrepancies are found.

7 Scramble all of the data by 2 Calculate the section BIP over
Exclusive-Oring with the the entire STS-1 and save for
2^7-1 PRBS. use in the next frame.

8 Insert the A1, A2, C1 bytes 1 Detect the frame word
into the SOH. Calculate the bytes to find the start
section BIP over the entire of the STS-1 frame.
STS-1, saving for use in the
next frame.

Figure 11 Assembling and disassembling the SONET signal


DS3 Payload Mapping

The SONET mapping that defines DS3 transport is asynchronous DS3. It is the least flexible SONET
mode, because DS3 is the lowest level which can be cross-connected without incurring the delay and
hardware cost of demultiplexing the SONET signal. Even though the mapping is less flexible than
mapping DS1 signals straight into SONET, the mapping exists primarily to transport the large amount
of DS3 which already exists in the network. Figure 12 shows the bit definitions of a single SPE row of
asynchronous DS3 mapping.

3 87 Bytes
Bytes

Section
Overhead

STS-1 Pointer STS-1 Synchronous
H1 H2 H3 Payload Envelope

POH R C I R C R I R C R O R S I
Line 8 18 1 205 16 2 6 208 16 2 2 2 1 1 208
9 Overhead
Bytes
STS-1 Path Overhead

Asynchronous
DS3 Mapping

Transport
Overhead

I = Information
POH = Path Overhead
R = Fixed Stuff
C = Control 87 Bytes
O = Data Link
S = Variable Stuff

Figure 12 Asynchronous DS3 mapping

The breakdown of each bit in the SPE row is as follows:
Information (I) = 621 bits/SPE row
Datalink (O) = 2 bits/SPE row
Control (C) = 5 bits/SPE row
Path Overhead (POH) = 8 bits/SPE row
Fixed Stuff (R) = 59 bits/SPE row
Variable Stuff (S) = 1 bit/SPE row
Total = 696 bits/SPE row


Currently, the datalink bits are undefined. Each row has the opportunity to use the variable stuff bit to
carry payload data, which is indicated by setting each of the five control bits to 1. Majority vote on the
control bits is used at the receiving end to detect if the variable stuff bit contains real or dummy
information, so that dummy bits can be removed from the signal before the DS3 is passed to other
asynchronous equipment, such as M13 multiplexers.
Using all of the variable stuff bits for data, it is possible to calculate the maximum DS3 rate which can
be transmitted with this mapping.
(1 variable bit/row + 621 data bits/row) x 9 rows/frame x 8,000 frames/second = 44.784 Mbps
Calculating the minimum DS3 rate which can be carried with this mapping is accomplished by
calculating the payload using none of the variable stuff bits.
621 data bits/row x 9 rows/frame x 8,000 frames/second = 44.712 Mbps
Since the nominal DS3 frequency is 44.736 Mbps, the average stuffing rate for this mapping can be
determined.
44.736 Mbps - 44.712 Mbps = 24 kbps of variable stuffing
The average stuff rate for a DS3 implies that three of the nine variable stuff bits are used in each frame
to carry data. As a side note, the total amount of overhead that is included with the SPE to transmit the
DS3 is calculated as follows:
(59 fixed stuff bits/row + 2 datalink bits/row + 5 control bits/row + 8 POH bits/row +
1 variable stuff bit/row) x 9 rows/frame x 8,000 frames/second = 5.4 Mbps

Virtual Tributaries

To transport payloads requiring less capacity than a DS3 signal, the 783-byte SPE is divided into seven
virtual tributary (VT) groups of 12 columns each. The seven groups are combined with the POH and
two fixed stuff columns to fill the entire STS-1 SPE.
VT Groups = 7 groups x 12 columns/ groups x 9 bytes/column = 756 bytes
Fixed Stuff = 2 columns x 9 bytes/column = 18 bytes
Path Overhead = 1 column x 9 bytes/column = 9 bytes
Total = 783 bytes
VT groups are analogous to DS2 framed signals. In other words, smaller tributaries can be multiplexed
and placed within a VT group. Individual VTs come in different sizes, termed VT1.5, VT2, VT3, and
VT6, to convey the approximate bandwidth which can be carried by the tributary, as shown in Figure
13. A VT1.5, for example, consumes three columns per STS-1 frame to accommodate the following bit
rate:
3 columns/frame x 9 bytes/column x 8 bits/byte x 8,000 frames/second = 1.728 Mbps
The VT1.5 is used to transport a DS1 at 1.544 Mbps plus required overhead. A VT2 uses four columns
per STS-1 frame, so its carrying capacity is:
4 columns/frame x 9 bytes/column x 8 bits/byte x 8,000 frames/second = 2.304 Mbps


3 Columns 6 Columns

1 1 2 3 1 1 2 3 4 5 6

2 4 7
VT1.5 VT3
3
9 9
Rows Rows
27 54
Bytes Bytes

27 27 54 54
125 µs 125 µs

4 Columns 12 Columns

1 1 2 3 4 1 1 2 3 4 5 6 7 8 9 10 11 12

2 5 2 13
VT2 VT6
3 3
9 9
Rows Rows
4 4
36 108
Bytes Bytes

36 36 108 108
125 µs 125 µs

Figure 13 Available virtual tributaries

Since there are 12 columns in a VT group, four individual VT1.5s may fit into a single VT group.
Likewise, only three VT2s, two VT3s, or one VT6 can fit in a group, as shown in Figure 14. Although
different VT groups within a single STS-1 SPE can carry different sized VTs, different sized VTs cannot
be combined within a single VT group. Figure 15 illustrates the STS-1 SPE configured to carry VT1.5s
in all seven VT groups, for a total of 28 VT1.5s.

4 3 2 1
9 VT 1.5 VT 2 VT 3 VT 6

12 12 12 12

Figure 14 VT group capacity


1 29 30 31 32 33 58 59 60 61 62 87

1 J1 R R

2 B3 R R

3 C2 R R

4 G1 R R

5 F2 1 2 3 28 R 1 2 3 28 R 1 2 3 28

6 H4 R R

7 Z3 R R

8 Z4 R R

9 Z5 R R

STS-1 Fixed Stuff Fixed Stuff
Path Overhead Bytes Bytes
Bytes

Figure 15 STS-1 frame configured to carry 28 VT1.5 payloads

VT1.5 Structure

For further study of VT structure, the asynchronous VT1.5 mapping will be used an an example. The
VT1.5 SPE is similar to the STS-1 SPE, since it contains dedicated performance monitoring overhead,
and a pointer is used to detect its start. However, structural differences make the VT SPE unique. A
VT1.5 SPE is divided over four consecutive STS-1 frames to form a superframe (SF) as shown in Figure
16. The VT overhead is directly analogous to the POH, since it travels with the DS1 from entry to exit
and contains additional end-to-end performance monitoring specific to the DS1. There are 771 bits in
the VT SF for data and two stuffing bits to compensate for timing differences caused by the
asynchronous DS1 payload. The breakdown of each bit in the VT1.5 SF is as follows:
Information (I) = 771 bits/SF
Stuff Opportunities (S) = 2 bits/SF
Stuff Control (C) = 6 bits/SF
Overhead (O) = 8 bits/SF
Fixed Stuff (R) = 13 bits/SF
VT Overhead = 40 bits/SF
Total = 840 bits/SF


Frame # Byte # VT1.5 Superframe

1 1 V1
VT Overhead Layer
1 2 V5
V1-V2 = Pointer Value
V3 = Pointer Action
1 3 R R R R R R I R V4 = Undefined
V5 = VT Status
J2 = VT Path Trace
Z6 = VT Path Growth
24 Bytes Z7 = VT Path Growth
(192 I Bits)
Bits
2 1 V2 I = Information
O = Overhead
C = Stuff Control
2 2 J2 S = Stuff Opportunity
R = Fixed Stuff
2 3 C1 C2 O O O O I R Z6 = VT Path Growth
Z7 = VT Path Growth

24 Bytes 108
(192 I Bits) Bytes

V3

Z6

C1 C2 O O O O I R

24 Bytes
(192 I Bits)

4 1 V4

4 2 Z7

4 3 C1 C2 R R R S1 S2 R

24 Bytes
(192 I Bits)

500 µs

Figure 16 VT1.5 superframe

Using all of the variable stuff bits in each SF for data, it is possible to calculate the maximum
asynchronous DS1 rate that can be transmitted with this mapping.
(2 stuff bit/SF + 771 data bits/SF) x 1 SF/4 frames x 8,000 frames/second = 1.546 Mbps
Calculating the minimum DS1 rate that can be carried with this mapping requires calculating the
payload using none of the variable stuff bits.
771 data bits/SF x 1 SF/4 frames x 8,000 frames/second = 1.542 Mbps
Since the nominal DS1 frequency is 1.544 Mbps, the average stuffing rate for this mapping can be
determined.
1.544 Mbps - 1.542 Mbps = 2 kbps of variable stuffing
A 2 kbps average stuff rate indicates an average use of one stuff bit per SF.


VT1.5 Mapping Modes

There are two conventional modes to map Vts into the SONET signal: locked and floating. The locked
mode uses fixed locations within the SPE for the VT data,allowing easy access to the 64 kbps voice
channel directly within the SONET signal. Although the STS-1 SPE is still allowed to float with respect
to the STS-1 frame in all mappings, the locked VT payload is not allowed to float with respect to the
VT overhead. This restriction prevents the VT cross-connects from adjusting the VT in the same
manner that is allowed at the SPE level. For this reason, the locked mode has been dropped altogether
from the ANSI T1.105 specification. Figure 17 illustrates the locked mapping.

3
Bytes 87 Bytes

Section
Overhead

Line
Overhead VT VT VT

9
Bytes
STS-1 Path Overhead

Locked-Mode Locked-Mode Locked-Mode
DS1 Mapping DS1 Mapping DS1 Mapping

Transport
Overhead

VT1.5 STS-1 Synchronous
Payload Envelope

Figure 17 Locked VT mapping

The prevailing multiplexing technique (floating mode) allows the lower-rate signal to retain minimum
timing consistency with the SONET network clock. This mapping permits the DS1 to float relative to
the VT overhead. Unlike locked mode, a floating VT uses a VT pointer to show the starting byte position
of the VT SPE within the VT payload structure. In this sense, the operation of the VT pointer is directly
analogous to the path pointer, and has the same advantages of minimizing payload buffers and
associated delay when cross-connecting at the VT level. Figure 18 shows conceptually how the path and
VT pointers are used to locate a particular VT payload in a SONET frame. The solid VT box combines
the V1-V3 bytes from the SF to represent the pointer. This pointer is incremented and decremented at
VT cross-connects in exactly the same manner as the path pointer, to compensate for timing differences
between two SONET signals.


3 87 Bytes
Bytes

Section
Overhead

STS-1 Pointer STS-1 Synchronous
H1 H2 H3 Payload Envelope

VT
Line
9 Overhead
VT Pointer
Bytes

STS-1 Path Overhead 1

2

3
125 µs

Transport VT Floating Mode
Overhead

Figure 18 Using the VT pointer

Floating mappings are further classified as channelized or unchannelized as an indication of the lowest
level of cross-connecting that can be accomplished with the mapping. The channelized (byte-
synchronous) mapping, as shown in Figure 19, is observable at the DS0 level and is consequently very
popular in integrated digital loop carrier (IDLC) applications. Since the DS0s can be removed and
inserted directly into the SONET signal without the typical cost and delay of an additional 1/0 cross-
connect, cost savings can be realized in the deployment of DS0 grooming architectures. This capability
technically would allow equipment to route calls through the local loop without having the call travel
through the main switch in the central office. When the billing software is able to support this
application, this technology will become more widespread. Additionally, the extension of SONET into
the local loop brings with it added protection and OAM&P messaging. The disadvantage of this
mapping is that it requires additional slip buffers to byte align the DS1 signal within the mapping, so it
may be slightly more expensive to implement than unchannelized systems.


3
Bytes 87 Bytes

Section
Overhead

Line
Overhead VT VT VT

9
Bytes

Byte-Synchronous
DS1 Mapping
STS-1 Path Overhead

Byte-Synchronous
DS1 Mapping
Transport
Overhead
Byte-Synchronous
DS1 Mapping

STS-1 Synchronous
Payload Envelope

VT1.5

VT1.5

Figure 19 Byte synchronous floating mode

Unchannelized (asynchronous and bit-synchronous) mappings are only observable at the VT level.
Asynchronous and bit-synchronous mappings are identical in physical appearance, as shown in Figure
20. The difference between these mappings is in the flexibility of the tributary stuffing as the DS1 enters
the SONET network, as described in VT1.5 Structure. Bit-synchronous systems are forced to use one
variable stuff bit per SF, while asynchronous mappings are allowed to compensate for differences
between the DS1 and the SONET clock. As a result, systems incorporating the bit-synchronous
mapping require the incoming DS1 to be timed directly to the SONET network clock. There are no slip
buffers required to implement these mappings, so equipment may be less expensive, making it popular
in long distance applications. For access to the DS0 channels, a 1/0 cross-connect is required for both
of these mappings, however transport systems are generally concerned with cross-connecting at higher
rates than voice. The efficiency and flexibility of the asynchronous mapping makes it the most
common.

Protection Switching

With upwards of 32 thousand telephone calls over a single OC-48, disaster recovery is substantially
more important to service providers today than it has been in the past. Therefore, a major requirement
of any future widely deployed transmission standard is the ability to withstand catastrophic failures
without severely affecting service. The SONET network is able to weather these failures due to
deployment of SONET ring architectures and automatic protection switching (APS) algorithms. These
mechanisms allow live traffic to flow through a new path whenever the old path is disrupted or becomes
degraded.


3 87 Bytes
Bytes

Section
Overhead

Line
Overhead VT VT VT
9
Bytes

Asynchronous or
Bit-Synchronous
DS1 Mapping
STS-1 Path Overhead

Asynchronous or
Bit-Synchronous
DS1 Mapping
Transport
Asynchronous or
Bit-Synchronous
DS1 Mapping

Overhead

STS-1 Synchronous
Payload Envelope

VT1.5

VT1.5

Figure 20 Asynchronous floating mode

Linear Systems

A linear (point-to-point) network can be implemented with working and protect (or primary and
secondary) fibers deployed in different locations (route diversity). Also called 1+1, both the working
and protect lines carry identical traffic, permitting the receiving end to monitor the status of each line
in real time. If the working line becomes degraded or is disrupted, the receiver simply switches to the
protect fiber. The degraded threshold is programmable and is usually set at a line BIP error rate of 1E-
7 or 8. This type of switching is also called “tail-switched,” since switching decisions take place at the
tail (receiving) end of a signal. Unfortunately the cost of fiber, receivers, and transmitters is doubled
between every protected node, as compared to a non-protected system. The cost of this protection can
be reduced by using a 1:n architecture, where n is between 1 and 14. This architecture is similar to 1+1
with two major differences. First, even in a 1:1 architecture, the protect line is not carrying the identical
traffic so the transmitting end must request a switch. Second, since there is one protect line for n
working lines, there is a possibility that a working line will not be granted a switch. The head
(transmitting) end determines the priority of the requestor and either honors or ignores the receiver’s
request to switch, hence this architecture is also called “head-switched.”


Unidirectional Path Switched Rings

The simplest ring is the 2-fiber unidirectional path switched ring (UPSR) as shown in Figure 21. The
term unidirectional is used to describe the direction of traffic under normal circumstances, or when the
ring is “clean”. In a UPSR, the traffic is only routed one direction (usually clockwise) unless troubles
occur. For example, traffic entering at point A and exiting at point B travels clockwise. Traffic entering
at point B and exiting at point A also travels clockwise.
Protection is accomplished by automatically bridging all traffic counterclockwise at entry nodes.
Exactly one-half of the capacity of the ring is therefore reserved for protection. On an OC-48 ring,
STS-1 channels 1 … 24 would be reserved for clockwise traffic, while channels 25 … 48 would be
reserved for counterclockwise protection. Protection channels can also be configured on a VT basis. The
node which handles the exit of the traffic simply selects the better of the two directions, much like 1+1
protection of a point-to-point system. The UPSR does not utilize automatic protection switching (APS)
messages for switching. The exit nodes examine individual path or VT overhead indicators to
independently select STS-1 or VT signals. Since no communication is required between the entry and
exit nodes, the protection switch time is not affected by the number of nodes in the ring, as is the case
with the line switched rings.

DS3 and DS1
Secondary

A
Primary

DS3 DS3
and and
DS1 B
DS1

DS3 and DS1

Figure 21 Unidirectional path switched ring

Bidirectional Line Switched Rings

Unlike UPSR, a bidirectional line switched ring (BLSR) may be architected with either 2-fibers or the
4-fiber ring as shown in Figure 22. A 2-fiber BLSR is similar to a UPSR, except that traffic is routed in
both directions around the ring under normal circumstances. For example, traffic entering at point A
and exiting at point B travels clockwise, while traffic entering at point B and exiting at point A travels
counterclockwise. This method allows the most efficient use of deployed equipment and fiber
resources.


DS3 and DS1
Secondary

A
Primary

DS3 DS3
and and
DS1 B
DS1

DS3 and DS1

Figure 22 Bidirectional line switched ring

In a BLSR, the STS-1 or VT traffic is not bridged in the opposite direction unless APS signaling between
the entry and the exit node specifically requests the channel be placed on protection. The APS
messaging may be accompanied by further instructions contained in the section DCC. Since the BLSR
requires communication, a switch time requirement of 50 milliseconds restricts the BLSR to 16 nodes.
An advantage is that bidirectional traffic allows network planners to periodically reroute signals for
purposes of load-balancing.
A 4-fiber BLSR uses two types of protection switching: span switching and ring switching. Normal
traffic is routed exactly as the 2-fiber BLSR, however if the transmit and receive fiber pair bundle is cut
or degraded between points A and B, a “span switch” occurs. The span switch routes traffic over the
protected fiber pair much like 1:1 protection on a point-to-point system, and no directional re-routing
is required. If both fiber pairs are degraded or a node fails, then a “ring switch” occurs by routing traffic
away from the failure. Unlike the UPSR, a BLSR examines only LOH performance indications to
determine quality of service. The monitoring does not extend to the path or VT layer.
Span switching and ring switching can be used simultaneously in a 4-fiber BLSR to protect traffic in the
event of multiple failures on the same ring. If both of these switching methods are used, the ring switch
will require “extra traffic” to be dropped if the ring is full. One advantage of the 4-fiber BLSR is that the
capacity is doubled over the 2-fiber rings since protection channels are not reserved, however the cost
of fiber optic cables, transmitters, and receivers is doubled.
One additional parameter which is used to describe protection switching mechanisms indicates what
happens when the original line has returned to an acceptable performance level. “Revertive” systems
will restore working traffic on the original path and “non-revertive” systems will simply change the
definition of “working” to describe the line which is currently being used.


Broadband Services

The SONET specification provides a means to offer services that require a larger bandwidth than a
single STS-1 (broadband) by uniting the previously independent STS-1s to form a phase and frequency
aligned pipe. An example is OC-3c, so named because it is formed by concatenating three STS-1s and
then transmitting them optically. The payload pointer in the first STS-1 points to the beginning of the
SPE as usual, but all three SPEs are aligned and referenced by this pointer to create a contiguous 149.760
Mbps envelope. Figure 23 illustrates a concatenated OC-3c payload envelope. The pointers in the
second and third STS-1 frames still physically exist along with a normal TOH, however they contain a
special value which indicates to use the pointer value from STS-1 Number 1. The traditional POH
columns in the SPE in the second and third STS-1 frames are replaced with data. Today, the demand
for concatenated pathways is very high to accommodate ATM growth.
OC-12c is another concatenated signal gaining in popularity, and provides 599.040 Mbps of capacity,
since there is one POH column for each group of three STS-1 frames throughout the OC-12. An
OC-12c can transport about 1.4 million ATM packets every second.

260

J1

B3

C2

G1

9 F2
Rows

H4 h h

Z3

Z4

Z5 h

STS-3c POH
53

h ATM Cell

Figure 23 STS-3c mapping for broadband services

Summary

The SONET signal embeds performance monitoring, maintenance, provisioning, and operations
information directly within the signal format. It combines mechanisms to allow for timing
inconsistencies throughout the network and provides a means for transporting a wide variety of
services. It allows tributary drop and insert while reducing equipment cost and timing delay to provide
on/off ramps to the industry analogy of the “superhighway.” Rings may be architected to provide a high
degree of service quality, even in the presence of multiple failure conditions. These advantages add up
to provide a powerful network standard which will continue to grow in popularity into this next era of
telecommunications improvements.


Appendix A: SONET Section, Line, Path and VT Overhead Layers

Many of the bytes in the tables of this appendix are undergoing further definition and/or modification
at the time of this writing.

Byte Name Description
SONET Section
A1-A2 Framing Bytes Provides frame alignment of each STS-1 within an STS-n (n = 1, 3, 12…).The value is hexadecimal F628.
J0-Z0 Trace Growth (STS-1 ID) Redefined byte to provide section trace functionally and growth
B1* Section BIP-8 Provides section error monitoring using a bit-interleaved parity 8 code (BIP-8) using even parity. It is calculated
over all bytes of the previous STS-n frame.
E1* Section Orderwire Provides a 64 kbps voice channel for communication between two STEs
F1 User 1 Reserved for user purposes
D1-D3* Section DCC Provides a 192 kbps Data Communications Channel (DCC) between two STEs, to allow for message-based adminis-
tration, monitoring, and other communications needs
*Only defined for the first STS-1 of an STS-n

SONET Line
H1-H2 Pointer Provides a byte offset value to indicate where the path overhead begins with each SPE
H3 Pointer Action Provides an extra byte for a negative stuff opportunity needed to perform a pointer decrement without losing any
data. It is defined for all STS-1s within an STS-n.
B2 Line BIP-8 Provides line error monitoring, by calculating a bit-interleaved, even parity check over all bits of the line overhead
and SPE, excluding the SOH of the previous STS-1 frame
K1-K2* APS Bytes Provides APS signaling between two LTEs
D4-D12* Line DOC Provides a 576 kbps DCC between two LTEs for administration, monitoring, and other communications
S1/Z1 Sync Status/Growth Provides information about the quality of the timing source. Also allows for future growth.
M0 REI-L/Growth Provides error count detected by LTE back to peer LTE
E2* Line Orderwire Provides a 64 kbps voice channel for communication between LTEs
*Only defined for the first STS-1 of an STS-n

SONET Path
J1 Path Trace Provides an indication of path connectivity by repeating a 64-byte fixed-length ACSII text string which is inserted
when the payload is mapped. Installation crews can modify the string to indicate the tributary source
B3 Path BIP-8 Provides path error monitoring, by calculating a bit-interleaved, even parity check over all bits of the previous SPE,
excluding the LOH and SOH
C2 Signal Label Provides an identification byte for the inserted payload
00 STS path unequipped
01 Equipped – non-specific payload
02 Floating VT mode
03 VT locked mode
04 Asynchronous mapping for DS3
12 Asynchronous mapping for DS4NA
13 Mapping for ATM
14 Mapping for DQDB
15 Asynchronous mapping for FDDI
E1-FC STS-1 payload with VT payload defects
G1 Path Status Provides a method for communicating for far-end path status back to the path originating equipment
F2 Path User Channel F2 is a 64 kbps channel reserved for user communication between two PTEs
H4 Multiframe Indicator Provides a multiframe phase indication of a VT payload to identify phases of a SF
Z3-Z5 Growth/User Partially reserved for growth and network provider layer information


SONET VT
V5 VT Status Provides error checking, signal label, and path status
J2 VT Path Trace Provides support for VT path trace function
Z6 VT Path Growth Future growth
Z7 VT Path Growth Future growth

All statements, technical information and recommendations related to the products herein are based upon informa-
tion believed to be reliable or accurate. However, the accuracy or completeness thereof is not guaranteed, and no
responsibility is assumed for any inaccuracies. The user assumes all risks and liability whatsoever in connection with
the use of a product or its application. JDSU reserves the right to change at any time without notice the design,
specifications, function, fit or form of its products described herein, including withdrawal at any time of a product
offered for sale herein. JDSU makes no representations that the products herein are free from any intellectual
property claims of others. Please contact JDSU for more information. JDSU and the JDSU logo are trademarks of
JDS Uniphase Corporation. Other trademarks are the property of their respective holders. ©2006 JDS Uniphase
Corporation. All rights reserved. 30137297 500 0106 FUNDSONET.TN.DAT.TM.AE

Test & Measurement Regional Sales

NORTH AMERICA LATIN AMERICA ASIA PACIFIC EMEA WEBSITE : www.jdsu.com
TEL : 1 866 228 3762 TEL : +55 11 5503 3800 TEL : +852 2892 0990 TEL : +49 7121 86 2222
FAX : +1 301 353 9216 FAX : +55 11 5505 1598 FAX : +852 2892 0770 FAX : +49 7121 86 1222

The fundamentals of sonet

More Related Content

What's hot (20)

Viewers also liked (8)

Similar to The fundamentals of sonet (20)

More from HARRY CHAN PUTRA (20)

Recently uploaded (20)

The fundamentals of sonet