Laboratory of Introduction to Communications (ICOM)

Lab Working Guide

Introducció a les Comunicacions (ICOM)

February, 2019

1
Laboratory of Introduction to Communications (ICOM)

Session 1. Basic operation of a spectrum analyzer based on

the swept superheterodyne receiver principle
Objective • Students will understand the basic behavior of
communication receivers, both fixed frequency
superheterodyne receivers and swept local oscillator
receivers.
• Students will be able to configure the spectrum analyzer to
properly analyze basic signals in the frequency domain.

Background • Students need to be familiar with general characteristics of

systems in communication receivers:
amplification/attenuation, frequency conversion, bandwidth,
envelope detection and noise.
• Students need concepts given in SIS of convolution, Fourier
transform, spectral content and modulation of signals, and
concepts of stochastic processes given in PPEE.

Manufacturer’s • R&S FSC Spectrum Analyzer Operating Manual (Chapter 4)

guide to consult in • Agilent Application Note 150. Spectrum Analysis Basics
the lab
(Available at Atenea)

Homework

Question 1.1 Read document “How Spectral Analyzer works” (4 pages) available at Atenea.

Question 1.2 Solve the following exercise and bring it to class.

Exercise. Let’s assume a signal generator with an internal impedance of 600Ω set to
provide a 500kHz sinusoidal signal with an open-circuit amplitude of 1.3V. Calculate the
power measured by a power meter with an internal impedance of 50Ω connected to this
signal generator. Express the result in dBm, where dBm is 10·log10[power in miliWatt]).

Activity in the lab:

Activity 1.1 Visualize the spectrum of a basic signal. Generate a 500kHz sinusoidal signal of
1.3V amplitude with the 600Ω signal generator with the help of the oscilloscope. Then,
connect it to the spectrum analyzer (SA) and observe its spectrum. Set the Reference Level,
Center Frequency and Span to properly visualize the generated signal, with some of the
higher order harmonics. Set the Display Range to 10dB per division. With the help of a
Marker verify that the power of the principal harmonic is equal to the result of Question 1.2.
Explain why the SA produces higher order harmonics.

Activity 1.2. Understanding information shown on the screen. Observe annotations shown
on the SA screen. Use FSCView program in the PC to capture the screen, copy it in your
report and explain the meaning of each acronym and annotation. Note that the Resolution

2
Laboratory of Introduction to Communications (ICOM)

and Video Filter Bandwidth, the input RF Attenuation and the Swept Time parameters are
automatically set by the SA once you enter the primary parameters (i.e. Center Frequency,
Span and Reference Level). However, you may modify them by setting the manual control
mode.

Activity 1.3. Shape of spectral components. The spectrum of the principal harmonic of the
sinusoid shown at the SA screen does not correspond to a Dirac delta function. Instead, a
wider shape located around 500kHz can be observed. To gain more insight into this aspect
of the SA, first adjust the configuration of the SA to visualize the spectrum shape around the
principal harmonic and set the Resolution Bandwidth control to manual mode. Then, change
the Resolution Bandwidth to 100kHz, observe the impact on the spectrum and compare this
value with the -3dB bandwidth of that shape (use the Display Line and/or Markers to
measure the bandwidth at -3dB). You may wish to get more familiar with this procedure by
trying other Resolution Bandwidth values. Finally, explain why the spectrum of the principal
harmonic adopts this particular shape.

Activity 1.4. Tuning with Zero Span. Set the Resolution Bandwidth control back to automatic
mode and deactivate Display Line. Set Span to 200kHz and, if needed, reconfigure the SA to
see the principal harmonic of the 500kHz signal at the center of the screen. Finally, select
Zero Span and observe the trace on the screen. What does this trace represent? What is the
relationship, if any, between that trace and the one you have for Span different from 0?

Activity 1.5. Impact of noise on weak signals and noise rejection procedures in the SA. Set
the Resolution Bandwidth control to automatic mode, and configure the SA to visualize the
principal and two first harmonics on the screen. Decrease the amplitude of the signal
generator until the power of the main spectral component is -50dBm (if necessary, pull out
the attenuation knob of the signal generator to further increase attenuation). Please note
that now the signal at the input of the SA is quite low (1mV on the impedance load) and the
harmonics are even lower. In your opinion, why is it so difficult to measure the frequency or
amplitude of these harmonics?

Two rejection noise procedures are available at the SA: Video filtering or Video Trace
Averaging. Video Filtering: Set the Video Filter Bandwidth control to manual mode and
reduce the Video Filter Bandwidth until a clean trace is observed so that the different
harmonics can be neatly identified. Measure the time required to generate the whole trace.
Video Trace Averaging: Set the Video Filter Bandwidth control to automatic mode, and
average a number of sweeps until a clean trace is observed on the screen (Trace Average in
the menu). Measure the time required to generate the whole trace. Discuss and compare
both procedures.

3
Laboratory of Introduction to Communications (ICOM)

Session 2. Characteristic parameters of a communications

receiver
Objective • Students will get familiar with the most relevant
characterization parameters of communication receivers.
• Students will learn techniques to measure these parameters.
• Students will learn about the interference caused by the
image frequency.

Background • Students need to be familiar with the configuration of the

Spectrum Analyzer and the procedure to make
measurements.
• Students need a basic knowledge of third order
intermodulation products at the output of the mixer.
• Students need to be familiar with the concepts of noise and
Signal-to-Noise Ratio (SNR).
• Students need to know the expression of a Gaussian function,
the concept of variance and a basic use of Matlab.

Manufacturer’s • Anritsu. Guide to Spectrum Analysis

guide to consult in • Agilent Application Note 150. Spectrum Analysis Basics
the lab
(Available at Atenea)

Homework

Question 2.1. Measurement of the Third Order Intercept characteristic point. Radiofrequency
(RF) communication receivers show a nonlinear effect when a high amplitude input signal
saturates the RF input mixer. This nonlinear behavior is characterized in communication
receivers by the so-called Third Order Intercept (TOI) point. TOI is defined as the power of two
equal-level tones at the input of the mixer so that the power of the third order intermodulation
products reaches the same level of the input tones. This parameter can be easily measured
driving the communication receiver with two tones of power 𝑃𝑃𝑖𝑖 and frequency f1 and f2, and
using

𝑇𝑇𝑇𝑇𝑇𝑇 (𝑑𝑑𝑑𝑑𝑑𝑑) = 𝑃𝑃𝑖𝑖 (𝑑𝑑𝑑𝑑𝑑𝑑) + ∆/2(𝑑𝑑𝑑𝑑) (1)

where ∆ is the ratio between 𝑃𝑃𝑖𝑖 and the power of the third order intermodulation products.

In this exercise we ask you to prove equation (1) using the following procedure. First assume
that the powers of the linear component at the output and at the input of the mixer are equal.
The power of the third order intermodulation products at the output of the mixer (that is, the
signals with frequencies 2𝑓𝑓1 ± 𝑓𝑓2 and 2𝑓𝑓2 ± 𝑓𝑓1 ) is proportional to the power of the input tones

4
Laboratory of Introduction to Communications (ICOM)

𝑃𝑃𝑖𝑖 cubed. Then express the power of the linear component and the power of the third order
intermodulation products in dBm, and show that the intersection point fulfills (1). Note that the
power of the third order intermodulation products versus the input power is a straight line with
a slope equal to three, whereas the power of the linear component versus the input power has
a unity slope.

Question 2.2. Measurement of Receiver Selectivity. The capability of a communication receiver

to reject adjacent channels is determined by the transition band of the Intermediate Frequency
(IF) filter. This feature is characterized by the Receiver Selectivity parameter, defined as the ratio
between the -60 dB bandwidth and the -3 dB bandwidth of the IF filter. Considering that the
Resolution filter in the SA corresponds to the IF filter, explain using your own words the
procedure you would follow to measure the Receiver Selectivity (-60dB vs. -3dB ratio) of the SA.
Please justify the input signal selected to properly visualize on the SA screen the transfer
function of the IF filter, and specify the configuration of the SA (Center Frequency, Span,
Resolution bandwidth, Reference Level and others) to measure the Receiver Selectivity.

Activity in the lab:

Activity 2.1. Measurement of Receiver Sensitivity. All receivers, including spectrum analysers,
generate some internal noise that limits its capability to deal with small signals. The Sensitivity
is the parameter that characterizes this feature, and is given by the noise floor in dBm of the
instrument for a particular IF filter bandwidth, usually the smallest Resolution Bandwidth
setting. In this activity we ask you to measure the sensitivity of your SA acting as a receiver of
10kHz bandwidth signals, with a carrier frequency between 500 and 1500 kHz and with a
minimum Signal-to-Noise Ratio (SNR) of 30 dB at the demodulator output. Follow the next
procedure to measure the Sensitivity.

• Configure the SA to operate with signals with a carrier frequency between 500 and 1500
kHz and 10kHz bandwidth. Set the attenuation to 0dB. Select your own carrier
frequency.
• Without connecting any signal to the receiver input, decrease the Reference Level until
you observe the noise floor.
• Measure the noise level setting Span to Zero and decrease the Video Filter bandwidth
until the trace is almost flat. You may use Display Line.

The Receiver Sensitivity for a 30dB of SNR would then be the noise floor level plus 30dB. Write
down the Receiver Sensitivity of your SA and justify the procedure followed.

Activity 2.2. Measurement of Third Order Intercept point. The objective of this activity is to
compute the TOI point of the SA using equation (1). For this purpose, generate two tones of
900kHz and 950kHz frequencies using the signal generators of 600Ω and 50Ω internal
impedance, respectively. Combine these signals using a power splitter specific for this exercise,
which consists in a resistive combining network matched to each generator’s internal
impedance. Apply the sum of these tones to the receiver RF input. Please note that the TOI value
is highly sensitive to the frequency range; therefore, the TOI measured in this exercise is valid
only for the AM range. Then, configure the receiver in swept mode (i.e. spectrum analyzer mode)

5
Laboratory of Introduction to Communications (ICOM)

with an Attenuation of 0dB, and select the Center Frequency and Span to visualize both the two
tones and the third order intermodulation products on the screen. Finally, adjust the level of the
two input tones to -20dBm.

At this stage, you might have to reduce the noise level that masks the third order
intermodulation products to properly measure the power level. If so, reduce the bandwidth of
the Video filter and the Resolution filter to set the noise level at least 60dB below the input tone
until the two intermodulation products are clearly visible.

Another aspect to consider before computing the TOI with equation (1) is to verify that the SA
is working well below saturation. Note that equation (1) is valid only for medium input power
values where the SA works below saturation since a compression gain effect is given otherwise.
To verify this, switch the RF Attenuator from 0 to 10 dB. Should you observe a change in the
power level of the input tones, it means that the SA is in saturation and you have to decrease
the input tones power well below -20dBm until this effect is not observed. Explain in your report
why this procedure is useful to verify whether the SA is working in saturation or not.

Compute the TOI point value using equation (1) and explain this activity in your report.

Activity 2.3. Measurement of the Receiver Selectivity. Following the procedure described in
Q2.2, measure the Selectivity of your SA with the Resolution Bandwidth set to 10kHz as the ratio
between the -50 dB bandwidth and the -3 dB bandwidth. Additionally, measure the bandwidth
of the Resolution filter at -5 dB, -10 dB, -20 dB and -30 dB. You may use the SA function “N dB
Down”.

The Resolution filter of the SA has a Gaussian transfer function as the following one.

Using the bandwidth measurements of the Resolution filter, adjust these measurements to a
Gaussian shape using the polyfit.m function of Matlab and determine the variance σ2. This
activity should be done at home and included in the report.

Question 2.3. Interference caused by image frequencies. Solve the following exercise on
interference caused by image frequencies at home and include it in the report.

Exercise. The figure shows a two-stage spectrum analyzer:

6
Laboratory of Introduction to Communications (ICOM)

The input signal is

with

=f1 1=
GHz, f 2 8.2428=
GHz, f 3 1.0428GHz
 f − f IF (1) 
H 1+ ( f ) =
Π  ; 2 B1 =
1GHz
 2 B 1 
 f − f IF (2) 
H 2+ ( f ) =
Π  ; 2 B2 =30 KHz
 2 B2 

Let’s assume the frequency of the first local oscillator is tuned to measure the power of
the input signal at frequency f1 .

a) Obtain the frequency of the first local oscillator.

b) Calculate the measured power in dBm, as a function of constants if the
spectrum analyzer input impedance is 50 . Please note that there are image
frequency terms contributing to the output power.

In the previous analysis we have considered both intermediate frequency filters with ideal
transfer functions. To evaluate the impact of a non ideal filter, let’s now consider the
following transfer functions

c) Obtain all the frequencies that will be present at the output of the second
intermediate frequency filter, create a table with a row for all the cosines coming from
the same input frequency of x(t) and complete it. It is not necessary to compute the
amplitude of each cosine at the output signal.
♦

7
Laboratory of Introduction to Communications (ICOM)

Session 3. Principles and operation of a Vector Signal

Analyzer
Objective • Students will understand the basic functions performed by
a vector signal analyzer (VSA), and the relation between
time and frequency parameters of the VSA .
• Students will be able to configure the VSA to visualize signals
properly.
• Students will use the VSA to study bandpass signals,
equivalent lowpass signals, and in-phase and quadrature
components.

Background • Students need to know the relation between time and

frequency parameters of discrete signals from the subject
SIS (1A).
• Students need to know the concept of signal windowing and
FFT.
• Students will use concepts of Lesson 3 of ICOM “Senyals I
Sistemes Passabanda”. Specifically, students need to know
how to compute the equivalent lowpass signal of a bandpass
signal.

Manufacturer’s guide • Keysight VSA 89600 User’s Guide

to consult in the lab • Online help of VSA 89600 software is of great help during
lab activities.

Homework

Question 3.1. Compute the equivalent lowpass signal, and the in-phase and quadrature
components of the following signal

choosing and as central frequency (a) , (b) and (c) .

Represent the Fourier transform of the equivalent lowpass signal.

Question 3.2. Read sections 1-7 of the Keysight VSA 89600 User’s Guide available at Atenea and
answer the short questions (a)-(e) below. Note that the user’s guide includes both hardware and
software aspects of the system. However, in the lab class you will work with Keysight VSA 89600
software to visualize recorded signals. You may also have a look at sections 11-14 to make
yourself more familiar with the software environment before starting the activity in the lab.

The system block diagram of section 4 shows how the signal is sampled and windowed before
computing the FFT, from which the VSA makes measurements or visualizes the spectrum.
Therefore, the VSA parameters frequency span, number of frequency points, resolution

8
Laboratory of Introduction to Communications (ICOM)

bandwidth, and time record length interrelate with each other. Sections 5-7 of the user’s guide
explain these relations, which will help you improve your understanding of the VSA operation.

a) Signal recording files can come either from radiofrequency (RF) or baseband signals
but they should be already in digital format. The file must include information about
the sampling frequency, the carrier (if any) and other data depending on the type of
signal. In the lab you will have to select the recording file to be visualized. Into which
stage of the system block diagram is the recording data fed?
b) What is the length of the time record (T) to achieve an effective resolution bandwidth
(RBW) of 10 Hz with a Hanning window?
c) What is the relation between the sampling frequency (Fs) and the resolution
bandwidth (RBW) of the spectrum of a RF signal calculated with the FFT in the VSA
89600? This relation shows that the resolution bandwidth can be improved either by
increasing the number of sample points (N) or by reducing the sampling frequency
(Fs). Which one is more efficient from the computational cost point of view? Justify
your answer.
d) Let’s assume we aim at visualizing the spectrum of a bandpass signal with a 10 kHz
bandwidth. If a Hanning window is used, select the number of sample points (N) to
achieve a resolution bandwidth of 100 Hz, when the sampling frequency (Fs) is chosen
using the Nyquist criteria.
e) Obtain the numerical relation between the frequency span (Span) and the sampling
frequency (Fs) in the VSA. Find the resolution bandwidth achieved with 500 sample
points of a 10kHz bandwidth signal, setting the Span to 50 kHz and using a Hanning
window.

Activity in the lab: (check the User’s Guide to complete the activities)

Activity 3.1. Acquisition and visualization of a recorded bandpass signal with Keysight VSA
89600. Start the VSA software and follow the procedure described in section 13 of Keysight VSA
89600 User’s Guide to analyze the recorded file “two_sinusoids_40dB.mat” (available at
directory “H:/Soft/Signals”). Please note that it is a .mat file. By default, the program displays 2
stacked subplots: the trace in the top subplot corresponds to the spectrum of the signal
recorded in the file, and the trace in the bottom subplot is the magnitude of the complex
envelope in logarithmic units. First of all, configure the display to show 3 stacked subplots with
the spectrum, the in-phase component of the signal (that is the real part of the equivalent
lowpass signal, denoted Real) and the quadrature component of the signal (i.e., the imaginary
part of the lowpass equivalent signal, denoted Imag). In order to select the trace to visualize in
each subplot first select the subplot A, B or C. Then select the type of data in the Trace/Data
menu, and finally go to Trace/Format to select the Real(I) or Imag(Q) signals. Stop the execution,
and answer these questions regarding the spectrum trace.
• Which is the maximum power level that can be represented in the figure?
• Which is the power ratio between two consecutive horizontal lines?
• Which are the minimum and the maximum frequencies represented in the figure?
• Determine the window used by the VSA using the values for the length of the time
record and for the resolution bandwidth shown on the screen.

9
Laboratory of Introduction to Communications (ICOM)

Activity 3.2. Measurements of frequency and magnitude. According to the spectrum, the
recorded signal corresponds to two sinusoids plus noise.

Measure the amplitude and the frequency of each sinusoid with the help of markers. The
procedure is as follows. First, in Pause mode place a marker at the first spectral component of
the spectrum using the Markers functions. The screen will show the frequency and power at the
marker’s position. Then, compute the amplitude of the sinusoid by converting the power in dBm
into a voltage in Volts, considering that the power is measured on an internal impedance of 50
Ω. Finally, repeat the process for the second spectral component and provide the temporal
expression of the recorded signal.

Activity 3.3. Visualization of the in-phase and quadrature components of a bandpass signal. In
question 3.1, you computed the equivalent lowpass signal of s(t) without noise, and its in-phase
and quadrature components for different central frequencies. In this activity we ask you to
compare these theoretical expressions with the practical ones plotted by the VSA. First of all,
change the central frequency of the VSA to f1 in the MeasSeatup/Frequency menu. Stop the
execution and compare the Real and Imag traces with the theoretical in-phase and quadrature
components you obtained in question 3.1 with . For this purpose, you may measure the
frequency and amplitude of the Real and Imag traces with the help of two markers (one of them
Normal and the other one Delta to provide differential measurements directly). Repeat twice
this exercise with a central frequency equal to f2 and with a central frequency equal to 1GHz,
and comment the results in your report.

Activity 3.4. Configuration of time and frequency parameters in the VSA. To modify the
parameters you may use the MeasSetup menu or you may click directly on the variable on the
screen.
a) Setting the central frequency equal to 1GHz, change the frequency Span to 36 MHz, 12
MHz, 4 MHz and 1 MHz and complete the following table. The number of frequency
points (M) is available at the MeasSetup/ResBW menu.

Span (MHz) Fs (MHz) M T (µs) RBW (KHz)

36

12

Discuss in your report the impact of Span on the rest of the parameters, taking into
account your answers to question 3.2 and sections 5-7 of the Keysight VSA 89600 User’s
Guide.

10
Laboratory of Introduction to Communications (ICOM)

Note that the temporal trace of the in-phase and quadrature components changes for
Span equal to 4 MHz and 1 MHz (they even vanish for 1 MHz). Therefore, it is important
to remark that the information shown on the screen corresponds to the signal visualized
in the spectrum in order to avoid aliasing.
b) With the frequency Span equal to 36MHz, increase the number of Frequency Points
gradually up to the maximum at the ResBW label. You will notice that by increasing the
number of Frequency Points not only does the trace denser in the frequency domain,
but also in the temporal domain. Explain in your report why this is observed.
c) Averaging either in the frequency domain or in the time domain can be useful to deal
with random signals. In our case, the signal has a noisy component that makes the
spectrum trace fluctuate. In this activity we ask you to measure the noise floor level
using the RMS (Video) function in the MeasSetup/Average menu. This function averages
a number of successive sweeps or instantaneous spectra as follows

𝑛𝑛
1
𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴 𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆(𝑓𝑓) = � � |𝐼𝐼𝐼𝐼𝐼𝐼𝐼𝐼𝐼𝐼𝐼𝐼𝐼𝐼𝐼𝐼𝐼𝐼𝐼𝐼𝐼𝐼𝐼𝐼𝐼𝐼 𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑖𝑖 (𝑓𝑓)|2
𝑛𝑛
𝑖𝑖=1

where i denotes the i-th realization of the spectrum and n is the number of realizations.
This average does not suppress the noise but estimates its mean power better at the
output of the resolution filter. Then, measure the noise floor level averaging 50
realizations of the spectrum at MeasSetup/Average. Do you observe any change in the
power of the signal at f1 and f2? Please note that the RMS (Video) function does not
affect the time domain traces at all.

11
Laboratory of Introduction to Communications (ICOM)

Session 4. Generation of digital signals with the LaVICAD

Simulator, and their analysis using a Vector Signal
Analyzer.
Objective
• Students will learn how to generate a QPSK digital signal with
the LaVICAD simulator. They will also make themselves familiar
with the square root raised cosine shaping pulse.
• Students will learn how to configure the VSA to operate as a
digital demodulator for the QPSK signal. With the help of the
VSA, they will visualize the signals at the output of the IQ
coherent digital demodulator, and will learn to interpret the
received signal constellation and I/Q eye diagrams.
• Students will better understand the impact of noise on the
received signal constellation, and on the I/Q eye diagrams. They
will learn the performance parameters of a digital demodulator
and how they can be measured with the help of the VSA.

Background
• Students should know the expression of the power spectral
density of a QPSK signal, and its equivalent lowpass signal.
• Students need to know the architecture of a digital QPSK
modulator and demodulator.
• Students will use concepts of Lesson 4 of ICOM “Modulacions
Digitals en canals AGWN”.

Manufacturer’s • Keysight VSA 89600 User’s Guide.

guide to consult in • Online help of Keysight VSA 89600.
the lab
(available at
Atenea or online
help)

Homework

Question 4.1. Obtain the power spectral density (PSD) of a QPSK (or 4 QAM) signal with a carrier
frequency of 10 MHz, and equiprobable bits at a rate of 64 kbit/s. Draw the resulting PSD
assuming a square root raised cosine (SRRC) pulse p(t) of unity energy and with a roll-off factor
of α=0.3. What is the bandwidth of this signal?

12
Laboratory of Introduction to Communications (ICOM)

Compare the PSD found with the one provided by the “QAM System” of the LaVICAD simulator
available at MatlabR2014b (see Activity 4.1). To obtain a graphical representation of the PSD,
you must configure the parameters of the first three stages of LaVICAD to generate the QPSK
signal defined above. Figure 5.1 shows a block diagram of the LaVICAD QAM Simulator. To
complete this exercise, just choose one of the available carrier frequencies.

Figure 5.1. Functional block diagram of LaVICAD QAM&PAM Simulator.

Question 4.2. The effective sampling frequency of the signal generated with LaVICAD depends
on two parameters: the number of samples per symbol to be selected in stage 2, and the symbol
rate in stage 3. On the other hand, the VSA 89600 requires a signal with a sampling frequency
that satisfies

Choose the values of the number of “samples/symbol” and “symbol rate” for the QPSK signal
defined in question 4.1 in order to visualize it properly with the VSA 89600 using a frequency
Span of 400 kHz.

Question 4.3. Draw a block diagram of a communication receiver of a QPSK digital signal that
uses a coherent IQ demodulator.

Activity in the lab:

In this session, you must generate a QPSK signal with the LaVICAD simulator, save it into a file
from the Lavicad main menu and configure the Vector Signal Analyzer, or Keysight VSA 89600
software, to operate as a digital communication receiver for this signal.

Activity 4.1. Generation of a QPSK digital signal with LaVICAD.

13
Laboratory of Introduction to Communications (ICOM)

To generate a digital signal with LaVICAD, first copy the folder H:/Soft/LavicadQAM_2018in your
desktop. Then, run MatlabR20014b and use the desktop as the current working directory.
Execute LavicadQAM from Matlab. Configure the LaVICAD simulator to generate a QPSK signal
using equiprobable bits at a rate of 64 Kbit/s, a square root raised cosine (SRRC) shaping pulse
and a carrier frequency of 10MHz. Use the results of Question 4.2 to select the number of
samples per symbol in stage 2 and the bit rate in stage 3. Write down in your report the values
you have selected for these parameters.
Then, select the option Adapt to AGILENT of the Session menu in LaVICAD to save the file that
can be uploaded with the VSA.

Activity 4.2. Visualization of the signal spectrum with the Vector Signal Analyzer.
In order to visualize the QPSK signal with the VSA, execute the VSA software and upload the file
saved in the previous activity. Remember that to upload a file to the VSA you may go to the
menu File/Recall/Recall Recording.
By default, one of the subplots should display the spectrum of the signal. Before proceeding,
verify that the center frequency is 10 MHz and that the frequency Span is 400 KHz. The objective
of this activity is that you compare this spectrum with the theoretical one obtained in Question
4.1. To do so, first average 50 realizations of the spectrum using the RMS (Video) function in the
MeasSetup/Average menu. Measure the bandwidth of the signal using the Marker OBW and
compare this result with the theoretical one. Comment the results in your report.

Activity 4.3. Configuration of the VSA as a demodulator.

First of all be sure the RMS (Video) function is deactivated. Then, activate the demodulator going
to MeasSetup/Measurement_Type:_Digital_Demod/General_Purpose:_Digital
Demod/DigitalDemod, and configure it through MeasSetup/Digital_Demod_Properties. You will
have to set up the following parameters in accordance with the QPSK signal generated before
(press Help button for comprehensive information):
a) Format is the modulation used by the signal.
b) Symbol Rate in number of symbols per second (i.e., 1 Hz = 1 baud)
c) Points/Symbol is the number of points used to depict one symbol in the time domain
(Volts vs. time) or in the modulation domain (Volt vs. Volt). Please note that this
parameter sets the resolution of the time traces but it has nothing to do with the
number of samples per symbol mentioned in Question 4.1
d) Result Length is the number of symbols to be used by the digital demodulator to show
graphical results on the screen such as eye diagrams or signal constellations, and to
compute values in the Symbol/Error table. It is important to take into account that
the Result Length value is also the number of symbols to be used by the carrier
recovery system. Therefore, the digital demodulator may have problems with the
carrier recovery function if Result Length is set too low.
e) Measurement Filter is the filter to be used in each branch of the IQ demodulator. It
should be the matched filter for the shaping pulse used in the incoming modulated
signal.
f) Reference Filter. One feature of the VSA consists in generating a reference signal that
may be compared with the signal at the output of the matched filter. This reference
signal is generated modulating the detected symbols with the pulse specified in

14
Laboratory of Introduction to Communications (ICOM)

Reference Filter. This filter should be equal to the autocorrelation of the shaping
pulse. Therefore, if the shaping pulse of the input signal is the square root raised
cosine, the measurement filter should be the raised cosine pulse.
g) Alpha/BT is the roll-off factor for SRRC pulses.
Propose a set of suitable values for these parameters considering that the input signal is a QPSK
signal with a SRRC pulse. Include them in your report and justify your choice.

Activity 4.4. Visualization of the Eye Diagram and Received Constellation.

Configure the display to show 4 subplots (grid 2x2) with the following traces: (A) Data IQ Meas
Time and Format Constellation, (B) Data Spectrum and Format Log Mag (dB), (C) Data IQ Meas
Time and Format Real (I), and (D) Data IQ Meas Time and Format Imag(Q).
The signal shown in subplot (C) is the output of the matched filter of the in-phase component
branch of the IQ demodulator. Similarly, the signal in (D) is the output of the matched filter of
the quadrature component branch of the IQ demodulator. Select a reduced number of symbols,
for instance 10 symbols, in each subplot to see the details of the signal properly (option X Scale
in the Trace menu). The vertical lines represent the time instants when the output of the
matched filter is sampled before the decision on the transmitted symbol is made.
Subplot (A) shows the constellation of the received signal. This figure represents into a two-
dimensional graph the samples at the output of the IQ demodulator, the horizontal axis
corresponds to the sample of the in phase component branch and the vertical axis to the
quadrature component branch. In order to gain more insight into this, do the following exercise.
Add one marker in subplots (C), (D) and (A) and activate the function “Couple Markers” in the
Markers menu. Explain in your report what happens when the markers are displaced. Do you
observe any dispersion in the constellation? Why?
Now, change subplot (D) to show the I-Eye format. The eye diagram is the superposition of
portions of the signal at the output of the matched filter of the in-phase component branch of
the IQ demodulator with a duration of 2 symbol periods (it can be similarly defined for the
output of the matched filter of the quadrature component branch of the IQ demodulator). A
measure of the performance of a digital communication system is the relative aperture of the
eye defined as d/D%, where d and D are shown in Figure 5.2. Using a larger number of symbols,
measure the relative aperture of the eye diagram for the in-phase component in the VSA and
explain the procedure in your report.

15
Laboratory of Introduction to Communications (ICOM)

Figure 5.2. Eye Diagram with partial closure.

Activity 4.5. Performance parameters of the digital demodulator when there are misalignment
errors.
The digital demodulator in Activity 4.4 performs perfectly because the digital demodulator is
configured properly and the signal has no noise. The objective of this activity is to study the
degradation observed in the demodulated signal when the demodulator is misaligned. For this
purpose, change the roll-off factor of the measurement filter to a significant different value from
the ideal one. Observe the dispersion in the constellation and the closure of the eye diagram.
Measure the relative aperture of the eye diagram of the in-phase component but using a much
larger number of symbols.

Activity 4.6. Performance parameters of the digital demodulator in the presence of additive
Gaussian noise.
In this activity we ask you to evaluate the degradation of digital demodulator performance in
the presence of noise.
Using LaVICAD generate the QPSK signal with a SRRC pulse of Activity 4.1 but now with an ideal
channel and additive Gaussian noise with an Eb/No of 15 dB. Modify the signal using Lavicad,
save the signal in a file, run the VSA and load the saved file. Configure the digital demodulator
to demodulate the new QPSK signal properly. Configure the display to show 4 subplots with the
following traces: (A) Data IQ Meas Time and Format Constellation, (B) Data Spectrum and
Format Log Mag (dB), (C) Data IQ Meas Time and Format I-Eye (I), and (D) Symbol/Errors.
Observe the closure of the eye diagram and the dispersion in the constellation due to the noise.
Comment these aspects in your report.
Subplot (D) shows the demodulator performance parameters of Error Vector Magnitude (EVM)
and Modulation Error rate (MER) that the VSA measures automatically. The EVM is the distance
between the received points in the constellation and the ideal one (see Figure 5.3 for a graphical
explanation).

Figure 5.3 Graphical representation of EVM(n) for the nth received symbol.

The VSA computes the EVM for each received symbol, and provides the root mean square (rms)
averaging and the maximum value (pk) in the Symbol/Error table both as a percentage of the
symbol energy. The MER is related to the received SNR and to the dispersion of the constellation.
Visit the online help for more information on the EVM and MER.

16
Laboratory of Introduction to Communications (ICOM)

The Syms/Errs table shows the maximum EVM, the rms of the EVM and the MER. It also indicates
the symbol that provides the maximum EVM(n). Using coupled markers in the constellation and
the Symbols/Errors table, observe that the symbol that maximizes the EVM is one which is far
away from the reference value. Generate new QPSK signal with different values of Eb/No, and
observe the impact on the EVM(rms) and on the MER.

17
Laboratory of Introduction to Communications (ICOM)

Session 5. Inter-Symbol Interference (ISI) and Baseband

Equalization
Objective • Students will better understand the effect of ISI on the
received signal space
• Students will have to design a linear equalizer.
Background • Students should know the concepts of BER and Eb/No.
• Students will use concepts in Lesson 5 of ICOM “Modulacions
digitals sobre canals limitats en banda”. Specifically, students
should know how to design an FIR digital linear equalizer using
the Zero Forcing criteria.
Manufacturer’s • Keysight VSA 89600 User’s Guide.
guide to consult in • Online help of Keysight VSA 89600.
the lab
(available at Atenea
or online help)

Homework

Question 5.1. Solve the exercise. A sequence of independent bits at 2Mbps bit rate with
modulates a QPSK signal with symbols given in the following
table

Symbol ai[·] aq[·]

00 1 1
01 -1 1
10 1 -1
11 -1 -1

and with a non-return to zero rectangular pulse of unit energy.

Consider the communication system in Figure 5.2, where is the channel impulse response,

w(t) is a zero-mean white Gaussian noise with spectral density equal to and
uncorrelated with the useful signal, and the lowpass filter of the coherent IQ demodulator is the
matched filter of the transmitted pulse, i.e. .

18
Laboratory of Introduction to Communications (ICOM)

Figure 5.1 Bandpass communication system

Assuming that the decision regions of the detector are the four quadrants of the received
constellation , answer the following questions.

a) Obtain the symbol rate, .

b) Considering W/Hz, obtain the Eb/No ratio in dB.
c) Assuming a channel , prove that the Bit Error Rate (BER) is equal to

where the Q-function is defined as

Note that in this case the equalizer is not needed since the channel is ideal.
Compute the BER for the Eb/No obtained in (b).

The equivalent lowpass system of Figure 5.1 is shown in Figure 5.2, where we have assumed the
IQ demodulator demodulates the bandpass signal without phase or frequency synchronization
errors. The numbers in red indicate the output of the corresponding LaVICAD stages.

Figure 5.2 Lowpass equivalent communication system of Figure 5.1.

Assume hereafter the following lowpass equivalent signal of the channel impulse response
. Answer the following questions.
d) Find the power spectral density of the received signal at stage (5).

19
Laboratory of Introduction to Communications (ICOM)

e) Obtain , i.e. the output of the matched filter .

f) Using the expression of the samples of the detected signal ,
identify the terms corresponding to the useful signal, the Inter-Symbol Interference
(ISI) and the noise. Plot the received constellation in the absence of
noise.
g) Calculate the resulting pulse at the output of the matched filter
shown in the figure and obtain the equivalent discrete impulse response of the
channel. Design the equalizer as a FIR filter with two coefficients, , under the
Zero Forcing criterion.
h) Calculate the samples of the resulting pulse at the output of the equalizer. Which is
the maximum value of ISI that will be produced for each sample? Calculate the noise
power at the output of the equalizer.

Question 5.2. Choose an appropriate value of the configuration parameters to simulate the
communication system of Figure 5.2 with the LaVICAD simulator considering a sampling
frequency equal to 16 MHz at stage (3). In particular, select the following parameters:

• Stage 1: Modulation format, bits/symbol, Symbol distance (mvolts)

• Stage 2: Pulse, Rolloff, Duration/T, Samples/Symbol
• Stage 3: Bit rate, Carrier Frequency
• Stage 4: Channel impulse response
• Stage 5: EbNo
• Stage 6: You can use values by default
• Stage 7: Equalizer type, Number of taps

Activity in the lab.

Activity 5.1. QPSK communication system with LaVICAD. Run LaVICAD from MATLAB and
simulate the exercise solved in Question 5.2 with a non-return to zero rectangular pulse
assuming an equivalent lowpass signal of the channel impulse response
. If you have not completed question (g) of the homeworks, use LaVICAD to generate a Zero
Forcing Equalizer with 3 coefficients. Observe the signal space at stages (6) and (7) using the
available options with and without noise. Comment the figures in your report.

Activity 5.2. Visualization of the transmitted QPSK spectrum with the VSA software. Visualize
with the VSA the QPSK transmitted signal, i.e, the signal generated at stage (3) of LaVICAD. Recall
the procedure to load files from LaVICAD to the VSA is explained in Session 4. Inside the VSA
software, average 50 realizations of the spectrum using the RMS (Video) function and visualize
the spectrum. Measure the symbol rate and the ratio in dB between the principal and secondary
lobe; compare it with the theoretical ones and comment the results in your report.

20
Laboratory of Introduction to Communications (ICOM)

Activity 5.3. Demodulation of the transmitted QPSK signal with VSA. Deactivate the RMS
(Video) function and activate the Digital Demodulator through the MeasSetup menu to
demodulate the transmitted QPSK signal. Recall this procedure is also explained in Session 4. Set
up all the configuration parameters, i.e. format, symbol rate, points/symbol, result length,
measurement filter, reference filter and alpha/BT. Then, configure the screen layout to have a
grid2x2 display with the traces of Spectrum (dB), IQ Constellation, I-Eye and Syms/Errs. Identify
the main differences between the eye diagram you observe now with a rectangular pulse and
the one you obtained in Session 4 where a square root raised cosine was used.

Activity 5.4. Visualization of the received QPSK spectrum without noise with VSA. Execute
again LaVICAD until stage (4) and visualize the received signal without noise with the VSA
software. Visualize the spectrum averaged with 50 realizations. What is the effect of the non-
ideal channel on the spectrum? Compare it to the theoretical shape calculated in Question 5.1.

Activity 5.5. Demodulation of the received QPSK signal without noise with VSA. Deactivate the
RMS (Video) function and activate the Digital Demodulator. Follow the steps as in Activity 5.3
and compare the results with the ones obtained there. Justify in your report the differences you
observe.

Activity 5.6. Demodulation of the received QPSK signal with noise with VSA. Load to VSA the
received signal with noise, i.e. the signal at the output of stage (5), with an Eb/No equal to the
one computed in Question 5.1. Activate the Digital Demodulator and visualize the spectrum and
the constellation. Measure the MER and compare it with the Eb/No used in the system. Open
Agilent help tool in order to check the MER definition if needed.

Activity 5.7. Design of the equalizer. Several criteria for designing equalizers can be found in the
literature. For instance, the Zero Forcing criterion, the MMSE criterion, and others more
complex as the one used by the VSA. You might be familiar with the Zero Forcing criterion if
already seen in ICOM lectures. In this activity we ask you to compare both the received
constellation (stage 7) and the probability of error (stage 8) obtained with LaVICAD without
equalizer and with the equalizer design in question 5.1 (g) If you have not completed question
(g) of the homeworks, use LaVICAD to generate a Zero Forcing Equalizer with 3 coefficients.
Comment the results in your report.

21