Imperfect waveforms measurement why and how

Author: Toppy Huang Page 1
IQXEL: WAVEFORM DEGRADATION WHY AND HOW?
Contents
1. A Brief Introduction of Direct Conversion and IQxel Waveform Degradation.....................................................2
2. IQxel Waveform Degradation Parameters ...........................................................................................................3
3. [Part 1]. DC Conversion.........................................................................................................................................4
3.1 IQ Gain Imbalance and Quadrature Skew Error..........................................................................................4
3.2 The effects of IQ imbalance and compensation on Receiver......................................................................5
***IQ Gain Imbalance Loopback Test on IQxel....................................................................................................5
3.3 Quadrature Skew Error (IQ Phase) Loopback Test on IQxel .......................................................................7
***IQ Phase Loopback Test on IQxel....................................................................................................................7
3.4 The Effects of DC Offset-LO Leakage...........................................................................................................8
***DC Offset Loopback Test on IQxel ..................................................................................................................9
3.5 The effects of Phase Noise and compensation on Receiver .....................................................................12
***Phase Noise Loopback Test on IQxel ............................................................................................................13
4. [Part2]. PA Non-Linear Effect..............................................................................................................................15
4.1 Backoff.......................................................................................................................................................15
***Backoff Loopback Test on IQxel....................................................................................................................15
4.2 SFactor-Smoothness Factor (non-linearity Rapp Model)..........................................................................16
***SFactor Loopback Test on IQxel....................................................................................................................16
5. [Part3]. Channel Effect........................................................................................................................................17
5.1 Channel Gain .............................................................................................................................................17
5.2 Channel Phase...........................................................................................................................................17
5.3 AWGN SNR ................................................................................................................................................18
***AWGN Loopback Test on IQxel.....................................................................................................................19
6. Courtesy..............................................................................................................................................................21

1. A Brief Introduction of Direct Conversion and IQxel Waveform Degradation
Compared to conventionally-used heterodyne architecture, Direct Conversion (zero IF,
homodyne-receivers) do not need to face the image problems in the I/Q arms. Therefore, bulky, off-chip,
front-end image-reject filter are unnecessary. However, the direct-conversion transceiver architecture
has some disadvantages such as DC offset via self-mixing, 1/f-noise, and the more severe IQ mismatch.
Nowadays, several compensation methods have been devised and implemented to the Wifi solutions.
The “Wave Degradation Settings” in IQxel GUI aims to let the users customize their imperfect waveforms
for developing and testing the robust radio systems.
In this document, we first introduce the principle for each parameter, and then perform the loopback
experiments to see how the impairments degrade the signals. All the impairments are generated based
on the same 11ac coding scheme as the following figure.
Figure 1 802.11ac BW20 MCS2 BCC waveform PSDU information
.

2. IQxel Waveform Degradation Parameters
 Direct Conversion
1. IQ Gain Imbalance
2. LO Phase Noise
3. I/Q DC Offset
 PA Non-linearity Effect
4. Backoff
5. SFactor
 Channel
6. Channel Gain
7. Channel Phase
8. AWGN SN
Figure 2 Common Sources of Error in a Direct Upconversion Transmitter

3. [Part 1]. DC Conversion
3.1 IQ Gain Imbalance and Quadrature Skew Error

3.2 The effects of IQ imbalance and compensation on Receiver
Figure 3 depicts a kth OFDM subcarrier and its mirror image, -k. The shape of the transmitted signal is
correct at the transmitter port (a), but the signal undergoes corruption on its way to the receiver due to
fading (b), and within the receiver due to IQ imbalance (c).
Figure 3 a) kth OFDM subcarrier and its mirror image –k at the transmitter port. b) Received signal at RX antenna port.
c) Direct-conversion received signal with IQ mismatch.
***IQ Gain Imbalance Loopback Test on IQxel
I/Q gain imbalance; specified per signal.
Valid values: - 6 to +6 in dB.
IQgain = Gain Inphase Signal Path ÷Gain Quadrature Signal Path
IQgain dB = 20log10(IQgain)
Figure 4 The effect of IQ gain imbalance of a QPSK signal

Figure 5 802.11ac SS1_MCS2_BCC original loopback test result without any degradations
Figure 6 802.11ac SS1_MCS2_BCC original loopback test result with IQ Gain Impairment 1dB

3.3 Quadrature Skew Error (IQ Phase) Loopback Test on IQxel
***IQ Phase Loopback Test on IQxel
I/Q phase imbalance in degrees; specified per signal. Valid values: - 30 to +30 in deg.
IQphase Imbalance is defined as:
IQphase = angle (Inphasesignal) − angle (Quadraturesignal ) − 90deg .

Figure 9 SS1_MCS2_BCC original loopback test result with IQ Phase 10 degree
Figure 10 SS1_MCS2_BCC original loopback test result with IQ Phase 30 degree
3.4 The Effects of DC Offset-LO Leakage
RF band selection is typically the only filtering performed in the receive chain of a direct conversion
receiver before the signal is down-converted directly to baseband. Therefore, a strong, nearby signal,
including the receiver's own LO, can mix with itself down to zero-IF (this is known as “self-mixing”) and
generate a dc level that appears as interference at the center of the desired band.

Figure 11 The effect of I and Q Offset on a QPSK signal
***DC Offset Loopback Test on IQxel
DC offset of I component in percent; specified per signal. DC offset in the I signal
path in the transmitter causes LO leakage.
Valid values: - 50 to +50 in percent.

Figure 12 802.11ac_BW20_SS1_MCS2_BCC without DC Offset
Figure 13 802.11ac_BW20_SS1_MCS2_BCC with 10% DC Offset

Figure 14 802.11ac_BW20_SS1_MCS2_BCC with 50% DC Offset

3.5 The effects of Phase Noise and compensation on Receiver
LO phase noise consists of contributions from the frequency stability of the reference crystal
oscillator, the frequency stability of the free-running voltage-controlled oscillator (VCO) used by the
phase-locked loop (PLL), and the loop bandwidth and noise from the PLL used in the frequency
synthesizer. The impact of phase noise can be seen as a circular distortion of the signal points
around the center of the symbol constellation diagram.
Ideally, a graph of power from a perfect oscillator as a function of frequency would be a single line,
meaning that the oscillator is generating its power by a single frequency.
However, because of noise on the power supply voltage, the oscillator frequency is actually being
frequency modulated over a very small range around the main oscillator frequency. This frequency
variation is called phase noise.
The oscillator output can be considered as a frequency spectrum.
Phase noise is defined as the power in a 1 Hz bandwidth at a frequency (fm) from the carrier. The
phase noise is measured in decibels below the carrier power (dBc).
Figure 15 Definition of Phase Noise
Figure 16 Phase Noise of PLO

***Phase Noise Loopback Test on IQxel
Figure 17 SS1_MCS2_BCC original loopback test result w.o. Phase Noise
Figure 18 SS1_MCS2_BCC original loopback test result w. Phase Noise= -70dBc/Hz

Figure 19 SS1_MCS2_BCC original loopback test result with Phase Noise= -50dBc/Hz

4. [Part2]. PA Non-Linear Effect
4.1 Backoff
In portable wireless device design, it is necessary to consider DC power consumption. This often requires
RF power amplifiers to operate near their compression point where the amplifier can achieve the highest
efficiency. As a result, the peak signal will experience some distortion due to nonlinear properties of the
amplifier.
Signal distortion due to amplifier nonlinearity will cause degradation of the EVM, and that distortion
point will determine the limit for the maximum transmitted power of a device. A WiFi signal typically has
peak to average ratio of about 10 dB.
***Backoff Loopback Test on IQxel
Power backoff from 1 dB compression point in dB; specified per signal. See also
smoothness factor setting below.
Valid values: 0 to 30 in dB.
Figure 20 802.11ac_BW20_SS1_MCS2_BCC with Backoff 0dB (CCDF shows a severe PA compression, EVM becomes worse, and the PSD
shows distortion)

4.2 SFactor-Smoothness Factor (non-linearity Rapp Model)
Figure 21 RAPP Model
***SFactor Loopback Test on IQxel
Smoothness factor used in Rapp model which is used to add compression; specified
per signal. See also Backoff setting above.
Valid values: 0 to 10.
Figure 22 802.11ac_BW20_SS1_MCS2_BCC with Backoff 0dB, SFactor=3 (EVM worse than P=2 )

5. [Part3]. Channel Effect
5.1 Channel Gain
Tx to Rx gain in dB. Set the gain in dB for channel matrix to be applied to the signals in specified
waveform. The dimension of the matrix is #VSG x #Tx Signals in the waveform. The matrix is input row by
row.
Valid values: -100 to +100 in dB
5.2 Channel Phase
Tx to Rx phase in degrees.
Valid values: -180 to +180 in degree

5.3 AWGN SNR
Additive white Gaussian noise (AWGN) is a channel model in which the only impairment to
communication is a linear addition of wideband or white noise with a constant spectral
density (expressed as watts per hertz of bandwidth) and a Gaussian distribution of amplitude. The model
does not account for fading, frequency selectivity, interference, nonlinearity or dispersion. However, it
produces simple and tractable mathematical models which are useful for gaining insight into the
underlying behavior of a system before these other phenomena are considered.
Wideband Gaussian noise comes from many natural sources, such as the thermal vibrations of atoms in
conductors (referred to as thermal noise or Johnson-Nyquist noise), shot noise, black body
radiation from the earth and other warm objects, and from celestial sources such as the Sun.
The AWGN channel is a good model for many satellite and deep space communication links. It is not a
good model for most terrestrial links because of multipath, terrain blocking, interference, etc. However,
for terrestrial path modeling, AWGN is commonly used to simulate background noise of the channel
under study, in addition to multipath, terrain blocking, interference, ground clutter and self interference
that modern radio systems encounter in terrestrial operation.

***AWGN Loopback Test on IQxel
Signal to noise ratio in dB which determines the amount of noise being added to signal; specified per
signal.
Valid values: -50 to 200 in dB.
Figure 23 802.11ac_BW20_SS1_MCS2_BCC with AWGN SNR=20dB (the PSD spillover and the constellation ideal points spread into a cloud)
Figure 24 802.11ac_BW20_SS1_MCS2_BCC with AWGN SNR=18dB (the PSD spillover and the constellation ideal points spread into a cloud)

Figure 25 PER_BER Curve
**SNR=Eb/N0+10log10(Data_rate/Bandwidth)
**PER = 1 - (1 - BER)^PL, where PL is lengh of the packet (header + payload).
For the waveform we used here, 802.11ac MCS2 (QPSK), 10% PER is approximately the same as a BER of
7e-5 (PL=1504 octets). When we set the SNR=16dB, i.e. Eb/N0=16.41dB, the demodulation is unable to
perform because BER=1e-4 at Eb/N0~16dB is equal to PER 33.6%!!!

6. Courtesy
1. Nutaq, “RF Imperfection and Compensation”
2. Angelfire, “Analogue IQ Error Correction for Transmitters-Off Line Method”
3. R,Svltek and S. Raman, “DC Offsets in Direct-Conversion Receivers: Characterization and Implications”
4. Evaluation Engineering, “Understanding WiMAX From the PHY Perspective”
5. Wiki, “AWGN”
6. Mark Webster, “Suggested PA Model for 802.11 HRb”
7. Allan W. Scott and Rex Frobeninus, RF Measurements for Cellular Phones and Wireless Data Systems

Imperfect waveforms measurement why and how

More Related Content

What's hot (19)

Viewers also liked (14)

Similar to Imperfect waveforms measurement why and how (20)

Imperfect waveforms measurement why and how