This page intentionally left blank

Software Receiver Design

Build Your Own Digital Communications System in Five Easy Steps
Have you ever wanted to know how modern digital communications systems work?
Find out with this step-by-step guide to building a complete digital radio that includes
every element of a typical, real-world communication system. Chapter by chapter, you
will create a MATLAB realization of the various pieces of the system, exploring the
key ideas along the way, as well as analyzing and assessing the performance of each
component. Then, in the nal chapters, you will discover how all the parts t together
and interact as you build the complete receiver. Can you decode the messages hidden
within the received signals?
In addition to coverage of crucial issues, such as timing, carrier recovery, and equalization, the text contains over 400 practical exercises, providing invaluable preparation for
industry, where wireless communications and software radio are becoming increasingly
important. Various extra resources are also provided online, including lecture slides and
a solutions manual for instructors.
C. Richard Johnson, Jr. is the Geoffrey S. M. Hedrick Senior Professor of Engineering
at Cornell University, where he has been on the faculty since 1981. He is a Fellow of the
IEEE and co-author of Telecommunication Breakdown (2004, with William A. Sethares)
and Theory and Design of Adaptive Filters (2001).
William A. Sethares is a Professor in the Department of Electrical and Computer Engineering at the University of Wisconsin in Madison. He is the author of Rhythm and
Transforms (2007) and Tuning, Timbre, Spectrum, Scale (2005).
Andrew G. Klein is an Assistant Professor at Worcester Polytechnic Institute. In addition
to working in academia, he has also held industry positions at several wireless start-up
companies.

Software Receiver Design

Build Your Own Digital Communications
System in Five Easy Steps
C. RICHARD JOHNSON, JR.
Cornell University

WILLIAM A. SETHARES
University of Wisconsin in Madison

ANDREW G. KLEIN
Worcester Polytechnic Institute

CAMBRIDGE UNIVERSITY PRESS

Cambridge, New York, Melbourne, Madrid, Cape Town,

Singapore, Sao Paulo, Delhi, Tokyo, Mexico City
Cambridge University Press
The Edinburgh Building, Cambridge CB2 8RU, UK
Published in the United States of America by Cambridge University Press, New York
www.cambridge.org
Information on this title: www.cambridge.org/9781107007529

C Cambridge University Press 2011
This publication is in copyright. Subject to statutory exception
and to the provisions of relevant collective licensing agreements,
no reproduction of any part may take place without the written
permission of Cambridge University Press.
First published 2011
Printed in the United Kingdom at the University Press, Cambridge
A catalogue record for this publication is available from the British Library
Library of Congress Cataloguing in Publication data
ISBN 978-1-107-00752-9 Hardback
ISBN 978-0-521-18944-6 Paperback
Additional resources for this publication at www.cambridge.org/9781107007529

Cambridge University Press has no responsibility for the persistence or

accuracy of URLs for external or third-party internet websites referred to in
this publication, and does not guarantee that any content on such websites is,
or will remain, accurate or appropriate.

To the Instructor . . .

. . . though its OK for the student to listen in.

Software Receiver Design helps the reader build a complete digital radio
that includes each part of a typical digital communication system. Chapter by
R
realization of the various pieces of the
chapter, the reader creates a Matlab
system, exploring the key ideas along the way. In the nal chapters, the reader
puts it all together to build fully functional receivers, though as Matlab code
they are not intended to operate in real time. Software Receiver Design
explores telecommunication systems from a very particular point of view: the
construction of a workable receiver. This viewpoint provides a sense of continuity
to the study of communication systems.
The three basic tasks in the creation of a working digital radio are
1. building the pieces,
2. assessing the performance of the pieces,
3. integrating the pieces.
In order to accomplish this in a single semester, we have had to strip away
some topics that are commonly covered in an introductory course and emphasize some topics that are often covered only supercially. We have chosen not to
present an encyclopedic catalog of every method that can be used to implement
each function of the receiver. For example, we focus on frequency division multiplexing rather than time or code division methods, and we concentrate only
on pulse amplitude modulation and quadrature amplitude modulation. On the
other hand, some topics (such as synchronization) loom large in digital receivers,
and we have devoted a correspondingly greater amount of space to these. Our
belief is that it is better to learn one complete system from start to nish than
to half-learn the properties of many.

Whole Lotta Radio

Our approach to building the components of the digital radio is consistent
throughout Software Receiver Design. For many of the tasks, we dene a
performance function and an algorithm that optimizes this function. This

vi

To the Instructor . . .

approach provides a unied framework for deriving the AGC, clock recovery, carrier recovery, and equalization algorithms. Fortunately, this can be accomplished
using only the mathematical tools that an electrical engineer (at the level of a
college junior) is likely to have, and Software Receiver Design requires no
more than knowledge of calculus, matrix algebra, and Fourier transforms. Any
of the ne texts cited for further reading in Section 3.8 would be ne.
Software Receiver Design emphasizes two ways to assess the behavior of
the components of a communication system: by studying the performance functions and by conducting experiments. The algorithms embodied in the various
components can be derived without making assumptions about details of the
constituent signals (such as Gaussian noise). The use of probability is limited to
naive ideas such as the notion of an average of a collection of numbers, rather
than requiring the machinery of stochastic processes. The absence of an advanced
probability prerequisite for Software Receiver Design makes it possible to
place it earlier in the curriculum.
The integration phase of the receiver design is accomplished in Chapters 9 and
15. Since any real digital radio operates in a highly complex environment, analytical models cannot hope to approach the real situation. Common practice
is to build a simulation and to run a series of experiments. Software Receiver
Design provides a set of guidelines (in Chapter 15) for a series of tests to verify
the operation of the receiver. The nal project challenges the digital radio that
the student has built by adding many dierent kinds of imperfections, including additive noise, multipath disturbances, phase jitter, frequency inaccuracies,
and clock errors. A successful design can operate even in the presence of such
distortions.
It should be clear that these choices distinguish Software Receiver Design
from other, more encyclopedic texts. We believe that this hands-on method
makes Software Receiver Design ideal for use as a learning tool, though it is
less comprehensive than a reference book. In addition, the instructor may nd
that the order of presentation of topics in the ve easy steps is dierent from
that used by other books. Section 1.3 provides an overview of the ow of topics,
and our reasons for structuring the course as we have.
Finally, we believe that Software Receiver Design may be of use to nontraditional students. Besides the many standard kinds of exercises, there are
many problems in the text that are self-checking in the sense that the reader
will know when/whether they have found the correct answer. These may also be
useful to the self-motivated design engineer who is using Software Receiver
Design to learn about digital radio.

To the Instructor . . .

vii

How Weve Used Software Receiver Design

The authors have taught from (various versions of) this text for a number of
years, exploring dierent ways to t coverage of digital radio into a standard
electrical engineering elective sequence.
Perhaps the simplest way is via a stand-alone course, one semester long, in
which the student works through the chapters and ends with the nal project
as outlined in Chapter 15. Students who have graduated tell us that when they
get to the workplace, where software-dened digital radio is increasingly important, the preparation of this course has been invaluable. After having completed
this course plus a rigorous course in probability, other students have reported
that they are well prepared for the typical introductory graduate-level class in
communications oered at research universities.
At Cornell University, the University of Wisconsin, and Worcester Polytechnic
Institute (the home institutions of the authors), there is a two-semester sequence
in communications available for advanced undergraduates. We have integrated
the text into this curriculum in three ways.
1. Teach from a traditional text for the rst semester and use Software
Receiver Design in the second.
2. Teach from Software Receiver Design in the rst semester and use a traditional text in the second.
3. Teach from Software Receiver Design in the rst semester and teach a
project-oriented extension in the second.
All three work well. When following the rst approach, students often comment
that by reading Software Receiver Design they nally understand what
they had been doing the previous semester. Because there is no probability
prerequisite for Software Receiver Design, the second approach can be moved
earlier in the curriculum. Of course, we encourage students to take probability
at the same time. In the third approach, the students were asked to extend the
basic pulse amplitude modulation (PAM) and quadrature amplitude modulation
(QAM) digital radios to incorporate code division multiplexing, to use more
advanced equalization techniques, etc.
We believe that the increasing market penetration of broadband communications is the driving force behind the continuing (re)design of radios (wireless
communications devices). Digital devices continue to penetrate the market formerly occupied by analog (for instance, digital television has now supplanted
analog television in the USA) and the area of digital and software-dened radio
is regularly reported in the mass media. Accordingly, it is easy for the instructor
to emphasize the social and economic aspects of the wireless revolution. The
impact of digital radio is vast, and it is an exciting time to get involved.

viii

To the Instructor . . .

Some Extras
The course website contains extra material of interest, especially to the instructor. First, we have assembled a complete collection of slides (in .pdf format)
that may help in lesson planning. The nal project is available in two complete
forms, one that exploits the block coding of Chapter 14 and one that does not.
In addition, there are several received signals on the website, which can be
used for assignments and for the project. Finally, all the Matlab code that is
presented in the text is available on the website. Once these are added to the
Matlab path, they can be used for assignments and for further exploration.1

Mathematical Prerequisites
r G. B. Thomas and R. L. Finney, Calculus and Analytic Geometry, 8th edition,
Addison-Wesley, 1992.

r B. Kolman and D. R. Hill, Elementary Linear Algebra, 8th edition, PrenticeHall, 2003.

r J. H. McClellan, R. W. Schafer, and M. A. Yoder, Signal Processing First,

Prentice-Hall, 2003.

The .m scripts will run with either Matlab or GNU Octave, which is freely available at
http://www.gnu.org/software/octave. When using the scripts with Matlab, the Signal Processing Toolbox is required; all scripts have been tested with Matlab v7.10/R2010a, but
are expected to work with older versions of Matlab. For Octave, the scripts were tested
with Octave v3.2.3 and the required Octave-Forge toolboxes signal v1.0.11, specfun v1.0.9,
optim v1.0.12, miscellaneous v1.0.9, and audio v1.1.4. When using Octave, the script
firpm octave.m can be renamed firpm.m so that identical code will run in the two platforms.

Contents

To the Instructor . . .

Step 1: The Big Picture

1

A Digital Radio
1.1 What Is a Digital Radio?
1.2 An Illustrative Design
1.3 Walk This Way

page v
1
2
2
3
12

Step 2: The Basic Components

15

A Telecommunication System
2.1 Electromagnetic Transmission of Analog Waveforms
2.2 Bandwidth
2.3 Upconversion at the Transmitter
2.4 Frequency Division Multiplexing
2.5 Filters that Remove Frequencies
2.6 Analog Downconversion
2.7 Analog Core of a Digital Communication System
2.8 Sampling at the Receiver
2.9 Digital Communications Around an Analog Core
2.10 Pulse Shaping
2.11 Synchronization: Good Times Bad Times
2.12 Equalization
2.13 Decisions and Error Measures
2.14 Coding and Decoding
2.15 A Telecommunication System
2.16 Stairway to Radio

16
16
18
20
22
23
24
26
28
29
30
33
34
35
37
38
38

The
3.1
3.2
3.3

40
41
44
46

Six Elements
Finding the Spectrum of a Signal
The First Element: Oscillators
The Second Element: Linear Filters

Contents

3.4
3.5
3.6
3.7
3.8

The Third Element: Samplers

The Fourth Element: Static Nonlinearities
The Fifth Element: Mixers
The Sixth Element: Adaptation
Summary

49
52
53
55
56

Step 3: The Idealized System

58

Modeling Corruption
4.1 When Bad Things Happen to Good Signals
4.2 Linear Systems: Linear Filters
4.3 The Delta Function
4.4 Convolution in Time: Its What Linear Systems Do
4.5 Convolution Multiplication
4.6 Improving SNR

59
59
65
65
70
72
76

Analog (De)modulation
5.1 Amplitude Modulation with Large Carrier
5.2 Amplitude Modulation with Suppressed Carrier
5.3 Quadrature Modulation
5.4 Injection to Intermediate Frequency

80
81
84
90
93

Sampling with Automatic Gain Control

6.1 Sampling and Aliasing
6.2 Downconversion via Sampling
6.3 Exploring Sampling in MATLAB
6.4 Interpolation and Reconstruction
6.5 Iteration and Optimization
6.6 An Example of Optimization: Polynomial Minimization
6.7 Automatic Gain Control
6.8 Using an AGC to Combat Fading
6.9 Summary

98
99
103
108
110
114
115
120
127
129

Digital Filtering and the DFT

7.1 Discrete Time and Discrete Frequency
7.2 Practical Filtering

130
130
141

Bits
8.1
8.2
8.3
8.4
8.5

152
152
155
157
160
161

to Symbols to Signals
Bits to Symbols
Symbols to Signals
Correlation
Receive Filtering: From Signals to Symbols
Frame Synchronization: From Symbols to Bits

Contents

Stu Happens
9.1 An Ideal Digital Communication System
9.2 Simulating the Ideal System
9.3 Flat Fading: A Simple Impairment and a Simple Fix
9.4 Other Impairments: More What Ifs
9.5 A B 3IG Deal

xi

165
166
167
175
178
187

Step 4: The Adaptive Components

191

10

Carrier Recovery
10.1 Phase and Frequency Estimation via an FFT
10.2 Squared Dierence Loop
10.3 The Phase-Locked Loop
10.4 The Costas Loop
10.5 Decision-Directed Phase Tracking
10.6 Frequency Tracking

192
194
197
202
206
210
216

11

Pulse Shaping and Receive Filtering

11.1 Spectrum of the Pulse: Spectrum of the Signal
11.2 Intersymbol Interference
11.3 Eye Diagrams
11.4 Nyquist Pulses
11.5 Matched Filtering
11.6 Matched Transmit and Receive Filters

226
227
229
231
237
242
247

12

Timing Recovery
12.1 The Problem of Timing Recovery
12.2 An Example
12.3 Decision-Directed Timing Recovery
12.4 Timing Recovery via Output Power Maximization
12.5 Two Examples

250
251
252
256
261
266

13

Linear Equalization
13.1 Multipath Interference
13.2 Trained Least-Squares Linear Equalization
13.3 An Adaptive Approach to Trained Equalization
13.4 Decision-Directed Linear Equalization
13.5 Dispersion-Minimizing Linear Equalization
13.6 Examples and Observations

270
272
273
284
288
290
294

14

Coding
14.1 What Is Information?
14.2 Redundancy

303
304
308

xii

Contents

14.3
14.4
14.5
14.6
14.7

Entropy
Channel Capacity
Source Coding
Channel Coding
Encoding a Compact Disc

315
318
323
328
339

Step 5: Putting It All Together

341

15

Make It So
15.1 How the Received Signal Is Constructed
15.2 A Design Methodology for the M6 Receiver
15.3 No Soap Radio: The M6 Receiver Design Challenge

342
343
345
354

16

A Digital Quadrature Amplitude Modulation Radio

16.1 The Song Remains the Same
16.2 Quadrature Amplitude Modulation (QAM)
16.3 Demodulating QAM
16.4 Carrier Recovery for QAM
16.5 Designing QAM Constellations
16.6 Timing Recovery for QAM
16.7 Baseband Derotation
16.8 Equalization for QAM
16.9 Alternative Receiver Architectures for QAM
16.10 The Q3 AM Prototype Receiver
16.11 Q3 AM Prototype Receiver Users Manual

357
357
358
363
367
378
380
384
387
391
397
398

Appendices

404

Transforms, Identities, and Formulas

A.1 Trigonometric Identities
A.2 Fourier Transforms and Properties
A.3 Energy and Power
A.4 Z-Transforms and Properties
A.5 Integral and Derivative Formulas
A.6 Matrix Algebra

404
404
405
409
409
410
411

Simulating Noise

412

Envelope of a Bandpass Signal

416

Relating the Fourier Transform to the DFT

D.1 The Fourier Transform and Its Inverse
D.2 The DFT and the Fourier Transform

421
421
422

Contents

Power Spectral Density

The
F.1
F.2
F.3
F.4

Averages and Averaging

G.1 Averages and Filters
G.2 Derivatives and Filters
G.3 Dierentiation Is a Technique, Approximation Is an Art

The
H.1
H.2
H.3

Z-Transform
Z-Transforms
Sketching the Frequency Response from the Z-Transform
Measuring Intersymbol Interference
Analysis of Loop Structures

B 3IG Transmitter
Constructing the Received Signal
Matlab Code for the Notorious B 3IG
Notes on Debugging and Signal Measurement

Index

xiii

425
428
428
432
435
438
442
442
443
446
451
453
455
459
460

xiv

Contents

With thanks to ...

Tom Robbins for his foresight, encouragement, and perseverance.
The professors who have taught from earlier versions of this book: Mark Andersland, Marc Buchner, Brian Evans, Edward Doering, Tom Fuja, Brian Sadler,
Michael Thompson, and Mark Yoder.
The classes of students who have learned from earlier versions of this book:
ECE436 and ECE437 at the University of Wisconsin, EE467 and EE468 at
Cornell University, and ECE 3311 at Worcester Polytechnic Institute.
Teaching assistants and graduate students who have helped rene this material: Jai Balkrishnan, Rick Brown, Wonzoo Chung, Qingxiong Deng, Matt
Gaubatz, Jarvis Haupt, Sean Leventhal, Yizheng Liao, Katie Orlicki, Jason
Pagnotta, Adam Pierce, Johnson Smith, Charles Tibenderana, and Nick
Xenias.
The companies that have provided data or access to engineers, and joined us
in the eld: Applied Signal Technology, Aware, Fox Digital, Iberium, Lucent
Technologies, and NxtWave Communications (now AMD).
The many talented academics and engineers who have helped hone this material: Raul Casas, Tom Endres, John Gubner, Rod Kennedy, Mike Larimore,
Rick Martin, Phil Schniter, John Treichler, John Walsh, Evans Wetmore, and
Robert Williamson.
The sta of talented editors at Cambridge University Press: Phil Meyler, Sarah
Finlay, Sarah Matthews, Steven Holt, and Joanna Endell-Cooper. Where
would we have been without LATEX guru Sehar Tahir?
Those with a well-developed sense of humor: Ann Bell and Betty Johnson.
Dread Zeppelin, whose 1999 performance in Palo Alto inspired the title Telecommunication Breakdown. In this new, improved, expanded (and more revealingly titled) version, the legacy continues: how many malapropic Led Zeppelin
song titles are there?

Dedicated to ...
Samantha and Franklin

Step 1: The Big Picture

Software Receiver Design: Build Your Own Digital Communications System

in Five Easy Steps is structured like a staircase with ve simple steps. The rst
chapter presents a naive digital communications system, a sketch of the digital
radio, as the rst step. The second chapter ascends one step to ll in details
and demystify various pieces of the design. Successive chapters then revisit the
same ideas, each step adding depth and precision. The rst functional (though
idealized) receiver appears in Chapter 9. Then the idealizing assumptions are
stripped away one by one throughout the remaining chapters, culminating in
sophisticated receiver designs in the nal chapters. Section 1.3 on page 12 outlines
the ve steps in the construction of the receiver and provides an overview of the
order in which topics are discussed.

A Digital Radio

1.1

What Is a Digital Radio?

The fundamental principles of telecommunications have remained much the same
since Shannons time. What has changed, and is continuing to change, is how
those principles are deployed in technology. One of the major ongoing changes is
the shift from hardware to softwareand Software Receiver Design reects
this trend by focusing on the design of a digital software-dened radio that you
will implement in Matlab.
Radio does not literally mean the AM/FM radio in your car; it represents
any through-the-air transmission such as television, cell phone, or wireless computer data, though many of the same ideas are also relevant to wired systems
such as modems, cable TV, and telephones. Software-dened means that key
elements of the radio are implemented in software. Taking a software-dened
approach mirrors the trend in modern receiver design in which more and more of
the system is designed and built in recongurable software, rather than in xed
hardware. The fundamental concepts behind the transmission are introduced,
demonstrated, and (we hope) understood through simulation. For example, when
talking about how to translate the frequency of a signal, the procedures are
presented mathematically in equations, pictorially in block diagrams, and then
concretely as short Matlab programs.
Our educational philosophy is that it is better to learn by doing: to motivate
study with experiments, to reinforce mathematics with simulated examples, to
integrate concepts by playing with the pieces of the system. Accordingly, each
of the later chapters is devoted to understanding one component of the transmission system, and each culminates in a series of tasks that ask you to build a
particular version of that part of the communication system. In the nal chapter,
the parts are combined to form a full receiver.
We try to present the essence of each system component in the simplest possible form. We do not intend to show all the most recent innovations (though our

A Digital Radio

presentation and viewpoint are modern), nor do we intend to provide a complete

analysis of the various methods. Rather, we ask you to investigate the performance of the subsystems, partly through analysis and partly using the software
code that you have created and that we have provided. We do oer insight into
all pieces of a complete transmission system. We present the major ideas of communications via a small number of unifying principles such as transforms to teach
modulation, and recursive techniques to teach synchronization and equalization.
We believe that these basic principles have application far beyond receiver design,
and so the time spent mastering them is well worth the eort.
Though far from optimal, the receiver that you will build contains all the
elements of a fully functional receiver. It provides a simple way to ask and answer
what if questions. What if there is noise in the system? What if the modulation
frequencies are not exactly as specied? What if there are errors in the received
digits? What if the data rate is not high enough? What if there are distortion,
reections, or echoes in the transmission channel? What if the receiver is moving?
The rst step begins with a sketch of a digital radio.

1.2

An Illustrative Design
The rst design is a brief tour of a digital radio. If some of the terminology
seems obscure or unfamiliar, rest assured that succeeding sections and chapters
will revisit the words and rene the ideas. The design is shown in Figures 1.1
through 1.5. While talking about these gures, it will become clear that some
ideas are being oversimplied. Eventually, it will be necessary to come back and
examine these more closely.
The boxed notes are reminders to
return and think about these areas
more deeply later on.
In keeping with Shannons goal of reproducing at one point a message known
at another point, suppose that it is desired to transmit a text message from one
place to another. Of course, there is nothing magical about text; however, .mp3
sound les, .jpg photos, snippets of speech, raster-scanned television images,
or any other kind of information would do, as long as it can be appropriately
digitized into ones and zeros.
Can every kind of message be digitized
into ones and zeros?
Perhaps the simplest possible scheme would be to transmit a pulse to represent
a one and to transmit nothing to represent a zero. With this scheme, however,
it is hard to tell the dierence between a string of zeros and no transmission at

Chapter 1. A Digital Radio

Text

Symbols
s[k]

Scaling
factor

Coder

Initiation
trigger

T-wide
analog
pulse
shape p(t)
generator

Baseband
signal y(t)

Figure 1.1 An idealized

baseband transmitter.

1
+ kT

all. A common remedy is to send a pulse with a positive amplitude to represent

a one and a pulse of the same shape but negative amplitude to represent a zero.
In fact, if the receiver could distinguish pulses of dierent sizes, then it would
be possible to send two bits with each symbol, for example, by associating the
amplitudes1 of +1, 1, +3, and 3 with the four choices 10, 01, 11, and 00.
The four symbols 1, 3 are called the alphabet, and the conversion from the
original message (the text) into the symbol alphabet is accomplished by the
coder in the transmitter diagram Figure 1.1. The rst few letters, the standard
ASCII (binary) representation of these letters, and their coding into symbols are
as follows
letter binary ASCII code
a
01 10 00 01
b
01 10 00 10
c
01 10 00 11
d
01 10 01 00
..
..
.
.

symbol string
1, 1, 3, 1
1, 1, 3, 1
1, 1, 3, 3
1, 1, 1, 3
..
.

(1.1)

In this example, the symbols are clustered into groups of four, and each cluster
is called a frame. Coding schemes can be designed to increase the security of a
transmission, to minimize the errors, or to maximize the rate at which data are
sent. This particular scheme is not optimized in any of these senses, but it is
convenient to use in simulation studies.
Some codes are better than others. How
can we tell?
To be concrete, let

r the symbol interval T be the time between successive symbols, and

r the pulse shape p(t) be the shape of the pulse that will be transmitted.
1

Many such choices are possible. These particular values were chosen because they are equidistant and so noise would be no more likely to ip a 3 into a 1 than to ip a 1 into a 1.

A Digital Radio

For instance, p(t) may be the rectangular pulse


1 when 0 t < T,
p(t) =
0 otherwise,

(1.2)

which is plotted in Figure 1.2(a). The transmitter of Figure 1.1 is designed so that
every T seconds it produces a copy of p() that is scaled by the symbol value s[].
A typical output of the transmitter in Figure 1.1 is illustrated in Figure 1.2(b)
using the rectangular pulse shape. Thus the rst pulse begins at some time
and it is scaled by s[0], producing s[0]p(t ). The second pulse begins at time
+ T and is scaled by s[1], resulting in s[1]p(t T ). The third pulse gives
s[2]p(t 2T ), and so on. The complete output of the transmitter is the sum
of all these scaled pulses:

s[i]p(t iT ).
y(t) =
i

Since each pulse ends before the next one begins, successive symbols should
not interfere with each other at the receiver. The general method of sending
information by scaling a pulse shape with the amplitude of the symbols is called
Pulse Amplitude Modulation (PAM). When there are four symbols as in (1.1),
it is called 4-PAM.
For now, assume that the path between the transmitter and receiver, which is
often called the channel, is ideal. This implies that the signal at the receiver is
the same as the transmitted signal, though it will inevitably be delayed (slightly)
due to the nite speed of the wave, and attenuated by the distance. When the
ideal channel has a gain g and a delay , the received version of the transmitted
signal in Figure 1.2(b) is as shown in Figure 1.2(c).
There are many ways in which a real signal may change as it passes from the
transmitter to the receiver through a real (nonideal) channel. It may be reected
from mountains or buildings. It may be diracted as it passes through the atmosphere. The waveform may smear in time so that successive pulses overlap. Other
signals may interfere additively (for instance, a radio station broadcasting at the
same frequency in a dierent city). Noises may enter and change the shape of
the waveform.
There are two compelling reasons to consider the telecommunication system
in the simplied (idealized) case before worrying about all the things that might
go wrong. First, at the heart of any working receiver is a structure that is able to
function in the ideal case. The classic approach to receiver design (and also the
approach of Software Receiver Design) is to build for the ideal case and later
to rene so that the receiver will still work when bad things happen. Second,
many of the basic ideas are clearer in the ideal case.
The job of the receiver is to take the received signal (such as that in Figure
1.2(c)) and to recover the original text message. This can be accomplished by an
idealized receiver such as that shown in Figure 1.3. The rst task this receiver
must accomplish is to sample the signal to turn it into computer-friendly digital

Chapter 1. A Digital Radio

y(t)
3
r(t)
p(t)

3g

g
0

T
1
(a)

+T

g
3g

+2T +3T
(b)

+ ++T
++2T ++3T
++4T
(c)

+4T

Figure 1.2 (a) An isolated rectangular pulse. (b) The transmitted signal consists of a

sequence of pulses, one corresponding to each symbol. Each pulse has the same shape
as in (a), though oset in time (by ) and scaled in magnitude (by the symbols s[k]).
(c) In the ideal case, the received signal is the same as the transmitted signal of (b),
though attenuated in magnitude (by g) and delayed in time (by ).
Reconstructed
symbols

Received
signal
Quantizer

Reconstructed
text
Decoder

Figure 1.3 An idealized

baseband receiver.

+ kT
Sampler

form. But when should the samples be taken? On comparing Figures 1.2(b) and
1.2(c), it is clear that if the received signal were sampled somewhere near the
middle of each rectangular pulse segment, then the quantizer could reproduce
the sequence of source symbols. This quantizer must either
1. know g so the sampled signal can be scaled by 1/g to recover the symbol
values, or
2. separate g from 3g and output symbol values 1 and 3.
Once the symbols have been reconstructed, then the original message can be
decoded by reversing the assignment of letters to symbols used at the transmitter
(for example, by reading (1.1) backwards). On the other hand, if the samples
were taken at the moment of transition from one symbol to another, then the
values might become confused.
To investigate the timing question more fully, let T be the sample interval and
be the time at which the rst pulse begins. Let be the time it takes for the signal
to move from the transmitter to the receiver. Thus the (k + 1)st pulse, which
begins at time + kT , arrives at the receiver at time + kT + . The midpoint
of the pulse, which is the best time to sample, occurs at + kT + + T /2. As
indicated in Figure 1.3, the receiver begins sampling at time , and then samples
regularly at + kT for all integers k. If were chosen so that
= + + T /2,

(1.3)

A Digital Radio

then all would be well. But there are two problems: the receiver does not know
when the transmission began, nor does it know how long it takes for the signal
to reach the receiver. Thus both and are unknown!
Somehow, the receiver must gure out
when to sample.
Basically, some extra synchronization procedure is needed in order to satisfy
(1.3). Fortunately, in the ideal case, it is not really necessary to sample exactly at
the midpoint; it is necessary only to avoid the edges. Even if the samples are not
taken at the center of each rectangular pulse, the transmitted symbol sequence
can still be recovered. But if the pulse shape were not a simple rectangle, then
the selection of would become more critical.
How does the pulse shape interact with
timing synchronization?
Just as no two clocks ever tell exactly the same time, no two independent
oscillators2 are ever exactly synchronized. Since the symbol period at the transmitter, call it Ttrans , is created by a separate oscillator from that creating the
symbol period at the receiver, call it Trec , they will inevitably dier. Thus another
aspect of timing synchronization that must ultimately be considered is how to
automatically adjust Trec so that it aligns with Ttrans .
Similarly, no clock ticks out each second exactly evenly. Inevitably, there is
some jitter, or wobble in the value of Ttrans and/or Trec . Again, it may be
necessary to adjust to retain sampling near the center of the pulse shape
as the clock times wiggle about. The timing adjustment mechanisms are not
explicitly indicated in the sampler box in Figure 1.3. For the present idealized
transmission system, the receiver sampler period and the symbol period of the
transmitter are assumed to be identical (both are called T in Figures 1.1 and
1.3) and the clocks are assumed to be free of jitter.
What about clock jitter?
Even under the idealized assumptions above, there is another kind of synchronization that is needed. Imagine joining a broadcast in progress, or one in
which the rst K symbols have been lost during acquisition. Even if the symbol
sequence is perfectly recovered after time K, the receiver would not know which
recovered symbol corresponds to the start of each frame. For example, using the
letters-to-symbol code of (1.1), each letter of the alphabet is translated into a
sequence of four symbols. If the start of the frame is o by even a single sym2

Oscillators, electronic components that generate repetitive signals, are discussed at length in
Chapter 3.

Chapter 1. A Digital Radio

bol, the translation from symbols back into letters will be scrambled. Does this
sequence represent a or X?
a




1, 1, 1, 3, 1
1, 1, 1, 3, 1



X

Thus proper decoding requires locating where the frame starts, a step called
frame synchronization. Frame synchronization is implicit in Figure 1.3 in the
choice of , which sets the time t (= with k = 0) of the rst symbol of the rst
(character) frame of the message of interest.
How to nd the start of a frame?
In the ideal situation, there must be no other signals occupying the same frequency range as the transmission. What bandwidth (what range of frequencies)
does the transmitter (1.1) require? Consider transmitting a single T -second-wide
rectangular pulse. Fourier transform theory shows that any such time-limited
pulse cannot be truly bandlimited, that is, cannot have its frequency content
restricted to a nite range. Indeed, the Fourier transform of a rectangular pulse
in time is a sinc function in frequency (see Equation (A.20) in Appendix A). The
magnitude of this sinc is overbounded by a function that decays as the inverse of
frequency (peek ahead to Figure 2.11). Thus, to accommodate this single-pulse
transmission, all other transmitters must have negligible energy below some factor of B = 1/T . For the sake of argument, suppose that a factor of 5 is safe, that
is, all other transmitters must have no signicant energy within 5B Hz. But this
is only for a single pulse. What happens when a sequence of T -spaced, T -wide
rectangular pulses of various amplitudes is transmitted? Fortunately, as will be
established in Section 11.1, the bandwidth requirements remain about the same,
at least for most messages.
What is the relation between the pulse
shape and the bandwidth?
One fundamental limitation to data transmission is the trade-o between the
data rate and the bandwidth. One obvious way to increase the rate at which data
are sent is to use shorter pulses, which pack more symbols into a shorter time.
This essentially reduces T . The cost is that this would require excluding other
transmitters from an even wider range of frequencies since reducing T increases
B.
What is the relation between the data
rate and the bandwidth?

A Digital Radio

If the safety factor of 5B is excessive, other pulse shapes that would decay
faster as a function of frequency could be used. For example, rounding the sharp
corners of a rectangular pulse reduces its high-frequency content. Similarly, if
other transmitters operated at high frequencies outside 5B Hz, it would be sensible to add a lowpass lter at the front end of the receiver. Rejecting frequencies
outside the protected 5B baseband turf also removes a bit of the higher-frequency
content of the rectangular pulse. The eect of this in the time domain is that the
received version of the rectangle would be wiggly near the edges. In both cases,
the timing of the samples becomes more critical as the received pulse deviates
further from rectangular.
One shortcoming of the telecommunication system embodied in the transmitter of Figure 1.1 and the receiver of Figure 1.3 is that only one such transmitter
at a time can operate in any particular geographical region, since it hogs all the
frequencies in the baseband, that is, all frequencies below 5B Hz. Fortunately,
there is a way to have multiple transmitters operating in the same region simultaneously. The trick is to translate the frequency content so that instead of all
transmitters trying to operate in the 0 and 5B Hz band, one might use the 5B
to 10B band, another the 10B to 15B band, etc. Conceivably, this could be
accomplished by selecting a dierent pulse shape (other than the rectangle) that
has no low-frequency content, but the most common approach is to modulate
(change frequency) by multiplying the pulse-shaped signal by a high-frequency
sinusoid. Such a radio-frequency (RF) transmitter is shown in Figure 1.4,
though it should be understood that the actual frequencies used may place it
in the television band or in the range of frequencies reserved for cell phones,
depending on the application.
At the receiver, the signal can be returned to its original frequency (demodulated) by multiplying by another high-frequency sinusoid (and then lowpass
ltering). These frequency translations are described in more detail in Section
2.6, where it is shown that the modulating sinusoid and the demodulating sinusoid must have the same frequencies and the same phases in order to return
the signal to its original form. Just as it is impossible to align any two clocks
exactly, it is also impossible to generate two independent sinusoids of exactly the
same frequency and phase. Hence there will ultimately need to be some kind of
carrier synchronization, a way of aligning these oscillators.
How can the frequencies and phases of
these two sinusoids be aligned?
Adding frequency translation to the transmitter and receiver of Figures 1.1 and
1.3 produces the transmitter in Figure 1.4 and the associated receiver in Figure
1.5. The new block in the transmitter is an analog component that eectively
adds the same value (in Hz) to the frequencies of all of the components of the
baseband pulse train. As noted, this can be achieved with multiplication by a
carrier sinusoid with a frequency equal to the desired translation. The new

10

Chapter 1. A Digital Radio

Text

Symbols
Coder

Pulseshape
filter

Received
Baseband
signal
signal
Frequency
Sampler
translator

Baseband
signal

Frequency
translator

Passband
signal

Reconstructed
text
Quantizer

Decoder

Figure 1.4

Radio-frequency
transmitter.

Figure 1.5

Radio-frequency
receiver.

block in the receiver of Figure 1.5 is an analog component that processes the
received analog signal prior to sampling in order to subtract the same value (in
Hz) from all components of the received signal. The output of this block should
be identical to the input to the sampler in Figure 1.3.
This process of translating the spectrum of the transmitted signal to higher
frequencies allows many transmitters to operate simultaneously in the same geographical area. But there is a price. Since the signals are not completely bandlimited to within their assigned 5B-wide slot, there is some inevitable overlap.
Thus the residual energy of one transmitter (the energy outside its designated
band) may interfere with other transmissions. Solving the problem of multiple
transmissions has thus violated one of the assumptions for an ideal transmission.
A common theme throughout Software Receiver Design is that a solution to
one problem often causes another!
There is no free lunch. How much does
the x cost?
In fact, there are many other ways in which the transmission channel can deviate from the ideal, and these will be discussed in detail later on (for instance, in
Section 4.1 and throughout Chapter 9). Typically, the cluttered electromagnetic
spectrum results in a variety of distortions and interferences:

r in-band (within the frequency band allocated to the user of interest)

r out-of-band (frequency components outside the allocated band such as the
signals of other transmitters)

r narrowband (spurious sinusoidal-like components)

r broadband (with components at frequencies across the allocated band and
r
r

beyond, including thermal noise introduced by the analog electronics in the

receiver)
fading (when the strength of the received signal uctuates)
multipath (when the environment contains many reective and absorptive
objects at dierent distances, the transmission delay will dier across dierent
paths, smearing the received signal and attenuating some frequencies more
than others)

A Digital Radio

Transmitted
signal

Gain
with
delay

Interference Broadband
noises
from other
sources
Received
signal
+
+

11

Figure 1.6 A channel model

admitting various kinds of

interferences.

Self-interference
Multipath

These channel imperfections are all incorporated in the channel model shown in
Figure 1.6, which sits in the communication system between Figures 1.4 and 1.5.
Many of these imperfections in the channel can be mitigated by clever use of
ltering at the receiver. Narrowband interference can be removed with a notch
lter that rejects frequency components in the narrow range of the interferer
without removing too much of the broadband signal. Out-of-band interference
and broadband noise can be reduced using a bandpass lter that suppresses
the signal in the out-of-band frequency range and passes the in-band frequency
components without distortion. With regard to Figure 1.5, it is reasonable to
wonder whether it is better to perform such ltering before or after the sampler
(i.e., by an analog or a digital lter). In modern receivers, the trend is to minimize
the amount of analog processing since digital methods are (often) cheaper and
(usually) more exible since they can be implemented as recongurable software
rather than xed hardware.
Analog or digital processing?
Conducting more of the processing digitally requires moving the sampler closer
to the antenna. The sampling theorem (discussed in Section 6.1) says that no
information is lost as long as the sampling occurs at a rate faster than twice the
highest frequency of the signal. Thus, if the signal has been modulated to (say)
the band from 20B to 25B Hz, then the sampler must be able to operate at least
as fast as 50B samples per second in order to be able to exactly reconstruct the
value of the signal at any arbitrary time instant. Assuming this is feasible, the
received analog signal can be sampled using a free-running sampler. Interpolation can be used to gure out values of the signal at any desired intermediate
instant, such as at time + kT (recall (1.3)) for a particular that is not an
integer multiple of T . Thus, the timing synchronization can be incorporated into
the post-sampler digital signal processing box, which is shown generically in
Figure 1.7. Observe that Figure 1.5 is one particular version of 1.7.
How exactly does interpolation work?

12

Chapter 1. A Digital Radio

Received
signal

Analog
signal
processing

+ kT

Digital
signal
processing

Recovered
source

Figure 1.7 A generic modern

receiver using both analog signal

processing (ASP) and digital
signal processing (DSP).

However, sometimes it is more cost eective to perform certain tasks in analog

circuitry. For example, if the transmitter modulates to a very high frequency, then
it may cost too much to sample fast enough. Currently, it is common practice
to perform some frequency translation and some out-of-band signal reduction in
the analog portion of the receiver. Sometimes the analog portion may translate
the received signal all the way back to baseband. Other times, the analog portion
translates to some intermediate frequency, and then the digital portion nishes
the translation. The advantage of this (seemingly redundant) approach is that
the analog part can be made crudely and, hence, cheaply. The digital processing
nishes the job, and simultaneously compensates for inaccuracies and aws in
the (inexpensive) analog circuits. Thus, the digital signal processing portion of
the receiver may need to correct for signal impairments arising in the analog
portion of the receiver as well as from impairments caused by the channel.
Use DSP when possible.
The digital signal processing portion of the receiver can perform the following
tasks:

r
r
r
r
r
r
r

downconvert the sampled signal to baseband

track any changes in the phase or frequency of the modulating sinusoid
adjust the symbol timing by interpolation
compensate for channel imperfections by ltering
convert modestly inaccurate recovered samples into symbols
perform frame synchronization via correlation
decode groups of symbols into message characters

A central task in Software Receiver Design is to elaborate on the system

structure in Figures 1.41.6 to create a working software-dened radio that can
perform these tasks. This concludes the illustrative design at the rst, most
supercial step of the radio stairway.
Use DSP to compensate for cheap ASP.

1.3

Walk This Way

This section provides a whirlwind tour of the complete structure of Software
Receiver Design. Each step presents the digital transmission system in greater
depth and detail.

A Digital Radio

13

r Step 1: The Naive Digital Communications System. As we have just seen,

the rst step introduced the digital transmission of data, and discussed how
bits of information may be coded into waveforms, sent across space to the
receiver, and then decoded back into bits. Since there is no universal clock,
issues of timing become important, and some of the most complex issues in
digital receiver design involve the synchronization of the received signal. The
system can be viewed as consisting of three parts:
1.
2.
3.

a transmitter as in Figure 1.4

a transmission channel
a receiver as in Figure 1.5

r Step 2: The Basic Components. The next two chapters provide more depth and

detail by outlining a complete telecommunication system. When the transmitted signal is passed through the air using electromagnetic waves, it must
take the form of a continuous (analog) waveform. A good way to understand
such analog signals is via the Fourier transform, and this is reviewed briey in
Chapter 2. The six basic elements of the receiver will be familiar to many readers, and they are presented in Chapter 3 in a form that will be directly useful
when creating Matlab implementations of the various parts of the communication system. By the end of the second step, the basic system architecture
is xed and the ordering of the blocks in the system diagram is stabilized.
Step 3: The Idealized System. The third step encompasses Chapters 4 through
9. Step 3 gives a closer look at the idealized receiverhow things work when
everything is just right: when the timing is known, when the clocks run at
exactly the right speed, when there are no reections, diractions, or diusions of the electromagnetic waves. This step also integrates ideas from previous systems courses, and introduces a few Matlab tools that are needed
to implement the digital radio. The order in which topics are discussed is
precisely the order in which they appear in the receiver:
frequency
channel translation sampling
Chapter 4
Chapter 5
Chapter 6
decision
receive
decoding
equalization
device
ltering







Chapter 7
Chapter 8
Channel impairments and linear systems Chapter
Frequency translation and modulation Chapter
Sampling and gain control
Chapter
Receive (digital) ltering
Chapter
Symbols to bits to signals
Chapter

4
5
6
7
8

14

Chapter 1. A Digital Radio

Chapter 9 provides a complete (though idealized) software-dened digital

radio system.
Step 4: The Adaptive Components. The fourth step describes all the practical
xes that are required in order to create a workable radio. One by one the
various problems are studied and solutions are proposed, implemented, and
tested. These include xes for additive noise, for timing oset problems, for
clock frequency mismatches and jitter, and for multipath reections. Again,
the order in which topics are discussed is the order in which they appear in
the receiver:
Carrier recovery
the timing of frequency translation
Receive ltering
the design of pulse shapes
Clock recovery
the timing of sampling
Equalization
lters that adapt to the channel
Coding
making data resilient to noise

Chapter 10
Chapter 11
Chapter 12
Chapter 13
Chapter 14

r Step 5: Putting It All Together. The nal steps are the projects of Chapters
15 and 16 which integrate all the xes of the fourth step into the receiver
structure of the third step to create a fully functional digital receiver. The wellfabricated receiver is robust with respect to distortions such as those caused
by noise, multipath interference, timing inaccuracies, and clock mismatches.

Step 2: The Basic Components

The next two chapters provide more depth and detail by outlining a complete
telecommunication system. When the transmitted signal is passed through the
air using electromagnetic waves, it must take the form of a continuous (analog)
waveform. A good way to understand such analog signals is via the Fourier
transform, and this is reviewed briey in Chapter 2. The six basic elements of
the receiver will be familiar to many readers, and they are presented in Chapter 3
in a form that will be directly useful when creating Matlab implementations of
the various parts of the communications system. By the end of the second step,
the basic system architecture is xed; the ordering of the blocks in the system
diagram has stabilized.

A Telecommunication System

Telecommunications technologies using electromagnetic transmission surround

us: television images icker, radios chatter, cell phones (and telephones) ring,
allowing us to see and hear each other anywhere on the planet. E-mail and the
Internet link us via our computers, and a large variety of common devices such
as CDs, DVDs, and hard disks augment the traditional pencil and paper storage
and transmittal of information. People have always wished to communicate over
long distances: to speak with someone in another country, to watch a distant
sporting event, to listen to music performed in another place or another time,
to send and receive data remotely using a personal computer. In order to implement these desires, a signal (a sound wave, a signal from a TV camera, or a
sequence of computer bits) needs to be encoded, stored, transmitted, received,
and decoded. Why? Consider the problem of voice or music transmission. Sending sound directly is futile because sound waves dissipate very quickly in air.
But if the sound is rst transformed into electromagnetic waves, then they can
be beamed over great distances very eciently. Similarly, the TV signal and
computer data can be transformed into electromagnetic waves.

2.1

Electromagnetic Transmission of Analog Waveforms

There are some experimental (physical) facts that cause transmission systems to
be constructed as they are. First, for ecient wireless broadcasting of electromagnetic energy, an antenna needs to be longer than about 1/10 of a wavelength
of the frequency being transmitted. The antenna at the receiver should also be
proportionally sized.
The wavelength and the frequency f of a sinusoid are inversely proportional. For an electrical signal travelling at the speed of light c (3 108 m/s),
the relationship between wavelength and frequency is
c
= .
f

A Telecommunication System

17

For instance, if the frequency of an electromagnetic wave is f = 10 kHz, then

the length of each wave is
=

3 108 m/s
= 3 104 m.
104 /s

Ecient transmission requires an antenna longer than 0.1, which is 3 km! Sinusoids in the speech band would require even larger antennas. Fortunately, there is
an alternative to building mammoth antennas. The frequencies in the signal can
be translated (shifted, upconverted, or modulated) to a much higher frequency
called the carrier frequency, at which the antenna requirements are easier to
meet. For instance,

r
r
r
r
r
r
r

AM radio: f 6001500 kHz 500200 m 0.1 > 20 m

VHF (TV): f 30300 MHz 101 m 0.1 > 0.1 m
UHF (TV): f 0.33 GHz 10.1 m 0.1 > 0.01 m
Cell phones (USA): f 824894 MHz 0.360.33 m 0.1 > 0.03 m
PCS: f 1.81.9 GHz 0.1670.158 m 0.1 > 0.015 m
GSM (Europe): f 890960 MHz 0.3370.313 m 0.1 > 0.03 m
LEO satellites: f 1.6 GHz 0.188 m 0.1 > 0.0188 m

Recall that 1 kHz = 103 Hz; 1 MHz = 106 Hz; 1 GHz = 109 Hz.
A second experimental fact is that electromagnetic waves in the atmosphere
exhibit dierent behaviors, depending on the frequency of the waves.

r Below 2 MHz, electromagnetic waves follow the contour of the Earth. This
r
r
r

is why shortwave (and other) radios can sometimes be heard hundreds or

thousands of miles from their source.
Between 2 and 30 MHz, sky-wave propagation occurs, with multiple bounces
from refractive atmospheric layers.
Above 30 MHz, line-of-sight propagation occurs, with straight-line travel
between two terrestrial towers or through the atmosphere to satellites.
Above 30 MHz, atmospheric scattering also occurs, which can be exploited
for long-distance terrestrial communication.

Humanmade media in wired systems also exhibit frequency-dependent behavior. In the phone system, due to its original goal of carrying voice signals, severe
attenuation occurs above 4 kHz.
The notion of frequency is central to the process of long-distance communications. Because of its role as a carrier (the AM/UHF/VHF/PCS bands mentioned
above) and its role in specifying the bandwidth (the range of frequencies occupied
by a given signal), it is important to have tools with which to easily measure the
frequency content in a signal. The tool of choice for this job is the Fourier trans-

18

Chapter 2. A Telecommunication System

form (and its discrete counterparts, the DFT and the FFT).1 Fourier transforms
are useful in assessing energy or power at particular frequencies. The Fourier
transform of a signal w(t) is dened as

W (f ) =
w(t)ej2f t dt = F {w(t)},
(2.1)
t=

where j = 1 and f is given in Hz (i.e., cycles/s or 1/s).

Speaking mathematically, W (f ) is a function of the frequency f . Thus, for
each f , W (f ) is a complex number and so can be plotted in several ways. For
instance, it is possible to plot the real part of W (f ) as a function of f and to plot
the imaginary part of W (f ) as a function of f . Alternatively, it is possible to plot
the real part of W (f ) versus the imaginary part of W (f ). The most common
plots of the Fourier transform of a signal are done in two parts: the rst graph
shows the magnitude |W (f )| versus f (this is called the magnitude spectrum)
and the second graph shows the phase angle of W (f ) versus f (this is called the
phase spectrum). Often, just the magnitude is plotted, though this inevitably
leaves out information.
Perhaps the best way to understand the Fourier transform is to look closely
at the inverse function

W (f )ej2f t df.
(2.2)
w(t) =
f =

The complex exponential ej2f t can be interpreted as a (complex-valued) sinusoidal wave since it is the sum of a sine term and a cosine term, both of frequency
f (via Eulers formula). Since W (f ) is a complex number at each f , (2.2) can
be interpreted as describing or decomposing w(t) into sinusoidal elements of frequencies f weighted by the W (f ). The discrete approximation to the Fourier
transform, called the DFT, is discussed in some detail in Chapter 7, and a table
of useful properties appears in Appendix A.

2.2

Bandwidth
If, at any particular frequency f0 , the magnitude spectrum is strictly positive
(|W (f0 )| > 0), then the frequency f0 is said to be present in w(t). The set of
all frequencies that are present in the signal is the frequency content, and if the
frequency content consists only of frequencies below some given f , then the
signal is said to be bandlimited to f . Some bandlimited signals are

r telephone-quality speech with maximum frequency 4 kHz and

r audible music with maximum frequency 20 kHz.
1

These are the discrete Fourier transform, which is a computer implementation of the Fourier
transform, and the fast Fourier transform, which is a slick, computationally ecient method
of calculating the DFT.

A Telecommunication System

19

|W( f )|
1
2
= 0.707
2

Figure 2.1 Various ways to

dene bandwidth (BW)

for a real-valued signal.

Half-power
BW
Null-to-null BW
Power BW
Absolute BW

But real-world signals are never completely bandlimited, and there is almost
always some energy at every frequency. Several alternative denitions of bandwidth are in common use; these try to capture the idea that most of the
energy is contained in a specied frequency region. Usually, these are applied
across positive frequencies, with the presumption that the underlying signals are
real-valued (and hence have symmetric spectra). Here are some of the alternative
denitions.
1. Absolute bandwidth is the smallest interval f2 f1 for which the spectrum is
zero outside of f1 < f < f2 (only the positive frequencies need be considered).
2. 3-dB (or half-power) bandwidth is f2 f1 , where, for frequencies outside f1 <
f < f2 , |H(f )| is never greater than 1/ 2 times its maximum value.
3. Null-to-null (or zero-crossing) bandwidth is f2 f1 , where f2 is the rst null
in |H(f )| above f0 and, for bandpass systems, f1 is the rst null in the envelope below f0 , where f0 is the frequency of maximum |H(f )|. For baseband
systems, f1 is usually zero.
4. Power bandwidth is f2 f1 , where f1 < f < f2 denes the frequency band in
which 99% of the total power resides. Occupied bandwidth is such that 0.5%
of power is above f2 and 0.5% below f1 .
These denitions are illustrated in Figure 2.1.
The various denitions of bandwidth refer directly to the frequency content
of a signal. Since the frequency response of a linear lter is the transform of its
impulse response, bandwidth is also used to talk about the frequency range over
which a linear system or lter operates.

20

Chapter 2. A Telecommunication System

Exercise 2.1. TRUE or FALSE: Absolute bandwidth is never less than 3-dB
power bandwidth.
Exercise 2.2. Suppose that a signal is complex-valued and hence has a spectrum
that is not symmetric about zero frequency. State versions of the various denitions of bandwidth that make sense in this situation. Illustrate your denitions
as in Figure 2.1.

2.3

Upconversion at the Transmitter

Suppose that the signal w(t) contains important information that must be transmitted. There are many kinds of operations that can be applied to w(t). Linear
time invariant (LTI) operations are those for which superposition applies, but
LTI operations cannot augment the frequency content of a signalno sine wave
can appear at the output of a linear operation if it was not already present in
the input.
Thus, the process of modulation (or upconversion), which requires a change
of frequencies, must be either nonlinear or time varying (or both). One useful
way to modulate is with multiplication; consider the product of the message
waveform w(t) with a cosine wave
s(t) = w(t) cos(2f0 t),

(2.3)

where f0 is called the carrier frequency. The Fourier transform can now be used
to show that this multiplication shifts all frequencies present in the message by
exactly f0 Hz.
Using one of Eulers identities (A.2),
cos(2f0 t) =

1 j2f0 t
e
+ ej2f0 t ,
2

(2.4)

one can calculate the spectrum (or frequency content) of the signal s(t) from the
denition of the Fourier transform given in (2.1). In complete detail, this is
S(f ) = F {s(t)} = F {w(t) cos(2f0 t)}




1 j2f0 t
e
+ ej2f0 t
= F w(t)
2




1 j2f0 t
w(t)
+ ej2f0 t ej2f t dt
=
e
2




1
=
w(t) ej2(f f0 )t + ej2(f +f0 )t dt
2


1
1
=
w(t)ej2(f f0 )t dt +
w(t)ej2(f +f0 )t dt
2
2
1
1
= W (f f0 ) + W (f + f0 ).
2
2

(2.5)

21

A Telecommunication System

|W( f )|

|S( f )|
0.5

f
(a)

f0 f

f0

f0 + f

f0 f

f0

f0 + f

(b)

Figure 2.2 Action of a modulator: If the message signal w(t) has the magnitude

spectrum shown in part (a), then the modulated signal s(t) has the magnitude
spectrum shown in part (b).

Thus, the spectrum of s(t) consists of two copies of the spectrum of w(t),
each shifted in frequency by f0 (one up and one down) and each half as large.
This is sometimes called the frequency-shifting property of the Fourier transform, and sometimes called the modulation property. Figure 2.2 shows how the
spectra relate. If w(t) has the magnitude spectrum shown in part (a) (this is
shown bandlimited to f and centered at 0 Hz or baseband, though it could be
elsewhere), then the magnitude spectrum of s(t) appears as in part (b). This
kind of modulation (or upconversion, or frequency shift) is ideal for translating
speech, music, or other low-frequency signals into much higher frequencies (for
instance, f0 might be in the AM or UHF bands) so that they can be transmitted eciently. It can also be used to convert a high-frequency signal back down
to baseband when needed, as will be discussed in Section 2.6 and in detail in
Chapter 5.
Any sine wave is characterized by three parameters: the amplitude, frequency,
and phase. Any of these characteristics can be used as the basis of a modulation
scheme: modulating the frequency is familiar from the FM radio, and phase
modulation is common in computer modems. A major example in this book is
amplitude modulation as in (2.3), where the message w(t) is multiplied by a highfrequency sinusoid with xed frequency and phase. Whatever the modulation
scheme used, the idea is the same: a sinusoid is used to translate the message
into a form suitable for transmission.
Exercise 2.3. Referring to Figure 2.2, nd which frequencies are present in
W (f ) and not in S(f ). Which frequencies are present in S(f ) but not in W (f )?
Exercise 2.4. Using (2.5), draw analogous pictures for the phase spectrum of
s(t) as it relates to the phase spectrum of w(t).
Exercise 2.5. Suppose that s(t) is modulated again, this time via multiplication
with a cosine of frequency f1 . What is the resulting magnitude spectrum? Hint:
let r(t) = s(t) cos(2f1 t), and apply (2.5) to nd R(f ).

22

Chapter 2. A Telecommunication System

2.4

Frequency Division Multiplexing

When a signal is modulated, the width (in Hz) of the replicas is the same as the
width (in Hz) of the original signal. This is a direct consequence of Equation (2.5).
For instance, if the message is bandlimited to f , and the carrier frequency is fc ,
then the modulated signal has energy in the range from f fc to +f fc and
from f + fc to +f + fc . If f  fc , then several messages can be transmitted
simultaneously by using dierent carrier frequencies.
This situation is depicted in Figure 2.3, where three dierent messages are
represented by the triangular, rectangular, and half-oval spectra, each bandlimited to f . Each of these is modulated by a dierent carrier (f1 , f2 , and f3 ),
which are chosen so that they are further apart than the width of the messages.
In general, as long as the carrier frequencies are separated by more than 2f ,
there will be no overlap in the spectrum of the combined signal. This process
of combining many dierent signals together is called multiplexing, and because
the frequencies are divided up among the users, the approach of Figure 2.3 is
called frequency-division multiplexing (FDM).
Whenever FDM is used, the receiver must separate the signal of interest from
all the other signals present. This can be accomplished with a bandpass lter as
in Figure 2.4, which shows a lter designed to isolate the middle user from the
others.
Exercise 2.6. Suppose that two carrier frequencies are separated by 1 kHz.
Draw the magnitude spectra if (a) the bandwidth of each message is 200 Hz
and (b) the bandwidth of each message is 2 kHz. Comment on the ability of the
bandpass lter at the receiver to separate the two signals.
Another kind of multiplexing is called time-division multiplexing (TDM), in
which two (or more) messages use the same carrier frequency but at alternating
time instants. More complex multiplexing schemes (such as code division multiplexing) overlap the messages in both time and frequency in such a way that
they can be demultiplexed eciently by appropriate ltering.
|W( f )|

f3 f * f3 + f *
Figure 2.3 Three dierent

f3

f2

f1

f
f1

f2
f2 f * f2 + f *

f3

upconverted signals are

assigned dierent frequency
bands. This is called
frequency-division
multiplexing.

A Telecommunication System

|W( f )|

23

Bandpass
filter
Figure 2.4 Separation

of a single FDM
transmission using a
bandpass lter.
f1

2.5

f2 f *

f2

f2 + f *

f3

Filters that Remove Frequencies

Each time the signal is modulated, an extra copy (or replica) of the spectrum
appears. When multiple modulations are needed (for instance, at the transmitter
to convert up to the carrier frequency, and at the receiver to convert back down to
the original frequency of the message), copies of the spectrum may proliferate.
There must be a way to remove extra copies in order to isolate the original
message. This is one of the things that linear lters do very well.
There are several ways of describing the action of a linear time-invariant lter.
In the time domain (the most common method of implementation), the lter is
characterized by its impulse response (which is dened to be the output of the
lter when the input is an impulse function). Because of the linearity, the output
of the lter in response to any arbitrary input is the superposition of weighted
copies of a time-shifted version of the impulse response, a procedure known as
convolution. Since convolution may be dicult to understand directly in the time
domain, the action of a linear lter is often described in the frequency domain.
Perhaps the most important property of the Fourier transform is the duality
between convolution and multiplication, which says that

r convolution in time multiplication in frequency, and

r multiplication in time convolution in frequency.
This is discussed in detail in Section 4.5. Thus, the convolution of a linear lter
can readily be viewed in the frequency (Fourier) domain as a point-by-point
multiplication. For instance, an ideal lowpass lter (LPF) passes all frequencies
below fl (which is called the cuto frequency). This is commonly plotted in a
curve called the frequency response of the lter, which describes the action of the
lter.2 If this lter is applied to a signal w(t), then all energy above fl is removed
from w(t). Figure 2.5 shows this pictorially. If w(t) has the magnitude spectrum
shown in part (a), and the frequency response of the lowpass lter with cuto
frequency fl is as shown in part (b), then the magnitude spectrum of the output
appears in part (c).
2

Formally, the frequency response can be calculated as the Fourier transform of the impulse
response of the lter.

24

Chapter 2. A Telecommunication System

|W( f )|

fl

(a)

fl

lter: (a) shows the magnitude

spectrum of the message which
is input into an ideal lowpass
lter with frequency response
(b); (c) shows the point-by-point
multiplication of (a) and (b),
which gives the spectrum of the
output of the lter.

|H( f )|

fl

(b)

fl

|Y( f )| = |H( f )| |W( f )|

fl

(c)

fl

Figure 2.5 Action of a lowpass

Exercise 2.7. An ideal highpass lter passes all frequencies above some given
fh and removes all frequencies below. Show the result of applying a highpass
lter to the signal in Figure 2.5 with fh = fl .

Exercise 2.8. An ideal bandpass lter passes all frequencies between an upper
limit f and a lower limit f. Show the result of applying a bandpass lter to the
signal in Figure 2.5 with f = 2fl /3 and f = fl /3.
The problem of how to design and implement such lters is considered in detail
in Chapter 7.

2.6

Analog Downconversion
Because transmitters typically modulate the message signal with a highfrequency carrier, the receiver must somehow remove the carrier from the message that it carries. One way is to multiply the received signal by a cosine wave
of the same frequency (and the same phase) as was used at the transmitter. This
creates a (scaled) copy of the original signal centered at zero frequency, plus some
other high-frequency replicas. A lowpass lter can then remove everything but
the scaled copy of the original message. This is how the box labelled frequency
translator in Figure 1.5 is typically implemented.
To see this procedure in detail, suppose that s(t) = w(t) cos(2f0 t) arrives at
the receiver, which multiplies s(t) by another cosine wave of exactly the same

A Telecommunication System

25

frequency and phase to get the demodulated signal

d(t) = s(t) cos(2f0 t) = w(t) cos2 (2f0 t).
Using the trigonometric identity (A.4), namely,
cos2 (x) =

1 1
+ cos(2x),
2 2

this can be rewritten as



1 1
1
1
d(t) = w(t)
+ cos(4f0 t) = w(t) + w(t) cos(2(2f0 )t).
2 2
2
2
The spectrum of the demodulated signal can be calculated to be


1
1
F {d(t)} = F
w(t) + w(t) cos(2(2f0 )t)
2
2
1
1
= F {w(t)} + F {w(t) cos(2(2f0 )t)}
2
2
by linearity. Now the frequency-shifting property (2.5) can be applied to show
that
1
1
1
(2.6)
F {d(t)} = W (f ) + W (f 2f0 ) + W (f + 2f0 ).
2
4
4
Thus, the spectrum of this downconverted received signal has the original baseband component (scaled to 50%) and two matching pieces (each scaled to 25%)
centered around twice the carrier frequency f0 and twice its negative. A lowpass
lter can now be used to extract W (f ), and hence to recover the original message
w(t).
This procedure is shown diagrammatically in Figure 2.6. The spectrum of the
original message is shown in (a), and the spectrum of the message modulated by
the carrier appears in (b). When downconversion is done as just described, the
demodulated signal d(t) has the spectrum shown in (c). Filtering by a lowpass
lter (as in part (c)) removes all but a scaled version of the message.
Now consider the FDM transmitted signal spectrum of Figure 2.3. This can be
demodulated/downconverted similarly. The frequency-shifting rule (2.5), with a
shift of f0 = f3 , ensures that the downconverted spectrum in Figure 2.7 matches
(2.6), and the lowpass lter removes all but the desired message from the downconverted signal.
This is the basic principle of a transmitter and receiver pair. But there are
some practical issues that arise. What happens if the oscillator at the receiver
is not completely accurate in either frequency or phase? The downconverted
received signal becomes s(t) cos(2(f0 + )t + ). This can have serious consequences for the demodulated message. What happens if one of the antennas is
moving? The Doppler eect suggests that this corresponds to a small nonzero
value of . What happens if the transmitter antenna wobbles due to the wind
over a range equivalent to several wavelengths of the transmitted signal? This
can alter . In eect, the baseband component is perturbed from (1/2)W (f ), and

26

Chapter 2. A Telecommunication System

|W( f )|
f

0
(a)

|S( f )|

f0

(b)

f0
Lowpass filter

{s(t) cos(2f0t)}

2f0

0
(c)

2f0

Figure 2.6 The message can be recovered by downconversion and lowpass ltering. (a)
shows the original spectrum of the message; (b) shows the message modulated by the
carrier f0 ; (c) shows the demodulated signal. Filtering with an LPF recovers the
original spectrum.

2f3

f3 f2 f3 f1

f1 f3

f2 f3

f3 f2

f3 f1

f3+f1

f3+f2

2f3

Figure 2.7 The signal containing the three messages of Figure 2.3 is modulated by a
sinusoid of frequency f3 . This translates all three spectra by f3 , placing two
identical semicircular spectra at the origin. These overlapping spectra, shown as
dashed lines, sum to form the larger solid semicircle. Applying an LPF isolates just
this one message.

simply lowpass ltering the downconverted signal results in distortion. Carrier

synchronization schemes (which attempt to identify and track the phase and frequency of the carrier) are routinely used in practice to counteract such problems.
These are discussed in detail in Chapters 5 and 10.

2.7

Analog Core of a Digital Communication System

The signal ow in the AM communication system described in the preceding
sections is shown in Figure 2.8. The message is upconverted (for ecient transmission), summed with other FDM users (for ecient use of the electromagnetic spectrum), subjected to possible channel noises (such as thermal noise),

A Telecommunication System

Other
FDM
Other
users interference
Received
signal

Message
waveform
Upconversion

Bandpass
filter for
FDM user
selection

Carrier
specification

Downconversion

27

Lowpass
filter

Carrier
synchronization

Figure 2.8 Analog AM communication system.

bandpass ltered (to extract the desired user), downconverted (requiring carrier
synchronization), and lowpass ltered (to recover the actual message).
But no transmission system operates perfectly. Each of the blocks in Figure 2.8
may be noisy, may have components which are inaccurate, and may be subject
to fundamental limitations. For instance,

r the bandwidth of a lter may be dierent from its specication (e.g., the
r
r
r
r

shoulders might not drop o fast enough to avoid passing some of the adjacent
signal),
the frequency of an oscillator might not be exact, and hence the modulation
and/or demodulation might not be exact,
the phase of the carrier is unknown at the receiver, since it depends on the
time of travel between the transmitter and the receiver,
perfect lters are impossible, even in principle,
no oscillator is perfectly regular, there is always some jitter in frequency.

Even within the frequency range of the message signal, the medium can
aect dierent frequencies in dierent ways. (These are called frequency selective
eects.) For example, a signal may arrive at the receiver, and a moment later a
copy of the same signal might arrive after having bounced o a mountain or a
nearby building. This is called multipath interference, and it can be viewed as
a sum of weighted and delayed versions of the transmitted signal. This may be
familiar to the (analog broadcast) TV viewer as ghosts, misty copies of the
original signal that are shifted and superimposed over the main image. In the
simple case of a sinusoid, a delay corresponds to a phase shift, making it more
dicult to reassemble the original message. A special lter called the equalizer is
often added to the receiver to help improve the situation. An equalizer is a kind
of deghosting circuit,3 and equalization is addressed in detail in Chapter 13.
3

We refrain from calling these ghost busters.

28

Chapter 2. A Telecommunication System

2.8

Sampling at the Receiver

Because of the proliferation of inexpensive and capable digital processors,
receivers often contain chips that are essentially special-purpose computers. In such receivers, many of the functions that are traditionally handled by discrete components (such as analog oscillators and lters) can
be handled digitally. Of course, this requires that the analog received signal be turned into digital information (a series of numbers) that a computer can process. This analog-to-digital conversion (A/D) is known as
sampling.
Sampling measures the amplitude of the waveform at regular intervals, and
then stores these measurements in memory. Two of the chief design issues in a
digital receiver are the following.

r Where should the signal be sampled?

r How often should the sampling be done?
The answers to these questions are intimately related to each other.
When taking samples of a signal, they must be taken fast enough that important information is not lost. Suppose that a signal has no frequency content above
f Hz. The widely known Nyquist reconstruction principle (see Section 6.1) says
that if sampling occurs at a rate greater than 2f samples per second, it is possible to reconstruct the original signal from the samples alone. Thus, as long as
the samples are taken rapidly enough, no information is lost. On the other hand,
when samples are taken too slowly, the signal cannot be reconstructed exactly
from the samples, and the resulting distortion is called aliasing.
Accordingly, in the receiver, it is necessary to sample at least twice as fast as
the highest frequency present in the analog signal being sampled in order to avoid
aliasing. Because the receiver contains modulators that change the frequencies
of the signals, dierent parts of the system have dierent highest frequencies.
Hence the answer to the question of how fast to sample is dependent on where
the samples will be taken.
The sampling
1. could be done at the input to the receiver at a rate proportional to the carrier
frequency,
2. could be done after the downconversion, at a rate proportional to the rate of
the symbols, or
3. could be done at some intermediate rate.
Each of these is appropriate in certain situations.
For the rst case, consider Figure 2.3, which shows the spectrum of the FDM
signal prior to downconversion. Let f3 + f be the frequency of the upper edge
of the user spectrum near the carrier at f3 . According to the Nyquist principle,
the upconverted received signal must be sampled at a rate of at least 2(f3 + f )
to avoid aliasing. For high-frequency carriers, this exceeds the rate of reasonably

A Telecommunication System

f3 fI

f3 + fI

f3 fI

29

f3 + fI

Figure 2.9 FDM downconversion to an intermediate frequency.

priced A/D samplers. Thus directly sampling the received signal (and performing
all the downconversion digitally) might not be feasible, even though it appears
desirable for a fully software-based receiver.
In the second case, the downconversion (and subsequent lowpass ltering)
are done in analog circuitry, and the samples are taken at the output of the
lowpass lter. Sampling can take place at a rate twice the highest frequency
f in the baseband, which is considerably smaller than twice f3 + f . Since the
downconversion must be done accurately in order to have the shifted spectra
of the desired user line up exactly (and overlap correctly), the analog circuitry
must be quite accurate. This, too, can be expensive.
In the third case, the downconversion is done in two steps: an analog circuit
downconverts to some intermediate frequency, where the signal is sampled. The
resulting signal is then digitally downconverted to baseband. The advantage of
this (seemingly redundant) method is that the analog downconversion can be
performed with minimal precision (and hence inexpensively), while the sampling
can be done at a reasonable rate (and hence inexpensively). In Figure 2.9, the
frequency fI of the intermediate downconversion is chosen to be large enough
so that the whole FDM band is moved below the upshifted portion. Also, fI is
chosen to be small enough so that the downshifted positive frequency portion
lower edge does not reach zero. An analog bandpass lter extracts the whole
FDM band at an intermediate frequency (IF), and then it is only necessary to
sample at a rate greater than 2(f3 + f fI ).
Downconversion to an intermediate frequency is common since the analog
circuitry can be xed, and the tuning (when the receiver chooses between users)
can be done digitally. This is advantageous since tunable analog circuitry is
considerably more expensive than tunable digital circuitry.

2.9

Digital Communications Around an Analog Core

The discussion so far in this chapter has concentrated on the classical core of
telecommunication systems: the transmission and reception of analog waveforms.
In digital systems, as considered in the previous chapter, the original signal
consists of a stream of data, and the goal is to send the data from one location
to another. The data may be a computer program, ASCII text, pixels of a picture,
a digitized MP3 le, or sampled speech from a cell phone. Data consist of a

30

Chapter 2. A Telecommunication System

sequence of numbers, which can always be converted to a sequence of zeros and

ones, called bits. How can a sequence of bits be transmitted?
The basic idea is that, since transmission media (such as air, phone lines, the
ocean) are analog, the bits are converted into an analog signal. Then this analog
signal can be transmitted exactly as before. Thus at the core of every digital
communication system lies an analog system. The output of the transmitter,
the transmission medium, and the front end of the receiver are necessarily analog.
Digital methods are not new. Morse-code telegraphy (which consists of a
sequence of dashes and dots coded into long and short tone bursts) became
widespread in the 1850s. The early telephone systems of the 1900s were analog,
with digitization beginning in the 1960s.
Digital (relative to fully analog) communications have the following advantages:

r
r
r
r
r
r
r
r

digital circuits are relatively inexpensive,

data encryption can be used to enhance privacy,
digital realization tends to support greater dynamic range,
signals from voice, video, and data sources can be merged for transmission
over a common system,
digital signals can be easily compressed
noise does not accumulate from repeater to repeater over long distances,
low error rates are possible, even with substantial noise,
errors can be corrected via coding.

In addition, digital receivers can easily be recongured or upgraded, because they

are essentially software-driven. For instance, a receiver built for one broadcast
standard (say for the American market) could be transformed into a receiver for
the European market with little additional hardware.
But there are also some disadvantages of digital (relative to fully analog)
communications, including the following:

r more bandwidth is (generally) required than with analog,

r synchronization is required.
2.10

Pulse Shaping
In order to transmit a digital data stream, it must be turned into an analog
signal. The rst step in this conversion is to clump the bits into symbols that
lend themselves to translation into analog form. For instance, a mapping from
the letters of the English alphabet into bits and then into the 4-PAM symbols
1, 3 was given explicitly in (1.1). This was converted into an analog waveform
using the rectangular pulse shape (1.2), which results in signals that look like

31

A Telecommunication System

(t)
1

Figure 2.10 The rectangular pulse

(t) of (2.8) is T time units wide

and centered at the origin.
T/2

T/2

Figure 1.2. In general, such signals can be written


s[k]p(t kT ),
y(t) =

(2.7)

where the s[k] are the values of the symbols, and the function p(t) is the pulse
shape. Thus, each member of the 4-PAM data sequence is multiplied by a pulse
that is nonzero over the appropriate time window. Adding all the scaled pulses
results in an analog waveform that can be upconverted and transmitted. If the
channel is perfect (distortionless and noise-free), then the transmitted signal will
arrive unchanged at the receiver. Is the rectangular pulse shape a good idea?
Unfortunately, though rectangular pulse shapes are easy to understand, they
can be a poor choice for a pulse shape because they spread substantial energy
into adjacent frequencies. This spreading complicates the packing of users in
frequency division multiplexing, and makes it more dicult to avoid having
dierent messages interfere with each other.
To see this, dene the rectangular pulse

1 T /2 t T /2
(t) =
(2.8)
0
otherwise
as shown in Figure 2.10. The shifted pulses (2.8) are sometimes easier to work
with than (1.2), and their magnitude spectra are the same by virtue of the timeshifting property (A.37). The Fourier transform can be calculated directly from
the denition (2.1)
T /2
 T /2

ej2f t 
j2f t
j2f t
(t)e
dt =
(1)e
dt =
W (f ) =
j2f t=T /2
t=
t=T /2
=

sin(f T )
ejf T ejf T
=T
T sinc(f T ).
j2f
f T

(2.9)

The sinc function is illustrated in Figure 2.11.

Thus, the Fourier transform of a rectangular pulse in the time domain is a
sinc function in the frequency domain. Since the sinc function dies away with
an envelope of 1/(x), the frequency content of the rectangular pulse shape
is (in principle) innite. It is not possible to separate messages into dierent
nonoverlapping frequency regions as is required for an FDM implementation as
in Figure 2.3.
Alternatives to the rectangular pulse are essential. Consider what is really
required of a pulse shape. The pulse is transmitted at time kT and again at time
(k + 1)T (and again at (k + 2)T . . .). The received signal is the sum of all these

32

Chapter 2. A Telecommunication System

1
Envelope
1/(x)
0.5

Figure 2.11 The sinc function

sinc(x) sin(x)/(x) has zeros at
every integer (except zero) and dies
away with an envelope of 1/(x).

sinc(x)

0
8

pulses (weighted by the message values). As long as each individual pulse is zero
at all integer multiples of T , then the value sampled at those times is just the
value of the original pulse (plus many additions of zero). The rectangular pulse
of width T seconds satises this criterion, as does any other pulse shape that
is exactly zero outside a window of width T . But many other pulse shapes also
satisfy this condition, without being identically zero outside a window of width
T.
In fact, Figure 2.11 shows one such pulse shapethe sinc function itself! It is
zero at all integers4 (except at zero, where it is unity). Hence, the sinc can be
used as a pulse shape. As in (2.7), the shifted pulse shape is multiplied by each
member of the data sequence, and then added together. If the channel is perfect
(distortionless and noise-free), the transmitted signal will arrive unchanged at
the receiver. The original data can be recovered from the received waveform by
sampling at exactly the right times. This is one reason why timing synchronization is so important in digital systems. Sampling at the wrong times may garble
the data.
To assess the usefulness of the sinc pulse shape, consider its transform. The
Fourier transform of the rectangular pulse shape in the time domain is the sinc
function in the frequency domain. Analogously, the Fourier transform of the sinc
function in the time domain is a rectangular pulse in the frequency domain (see
(A.22)). Thus, the spectrum of the sinc is bandlimited, and so it is appropriate
for situations requiring bandlimited messages, such as FDM. Unfortunately, the
sinc is not entirely practical because it is doubly innite in time. In any real
implementation, it will need to be truncated.
The rectangular and the sinc pulse shapes give two extremes. Practical pulse
shapes compromise between a small amount of out-of-band content (in frequency) and an impulse response that falls o rapidly enough to allow reasonable truncation (in the time domain). Commonly used pulse shapes such as the
square-root raised cosine shape are described in detail in Chapter 11.
4

In other applications, it may be desirable to have the zero crossings occur at places other
than the integers. This can be done by suitably scaling the x.

A Telecommunication System

p1(t)

p2(t)

p3(t)

T/2
1

T/2
1

T/2

33

Figure 2.12 Some

alternative pulse
shapes.

Exercise 2.9. Consider the three pulse shapes sketched in Figure 2.12 for a
T -spaced PAM system.
a. Which of the three pulse shapes in Figure 2.12 has the largest baseband power
bandwidth? Justify your answer.
b. Which of the three pulse shapes in Figure 2.12 has the smallest baseband
power bandwidth? Justify your answer.

Exercise 2.10. TRUE or FALSE: The atter the top of the pulse shape, the
less sensitive the receiver is to small timing osets. Explain your reasoning.

2.11

Synchronization: Good Times Bad Times

Synchronization may occur in several places in the digital receiver.

r Symbol phase synchronizationchoosing when (within each interval T ) to

sample.

r Symbol frequency synchronizationaccounting for dierent clock (oscillator)

rates at the transmitter and receiver.

r Carrier phase synchronizationaligning the phase of the carrier at the receiver

with the phase of the carrier at the transmitter.

r Carrier frequency synchronizationaligning the frequency of the carrier at

the receiver with the frequency of the carrier at the transmitter.

r Frame synchronizationnding the start of each message data block.

In digital receivers, it is important to sample the received signal at the appropriate time instants. Moreover, these time instants are not known beforehand;
rather, they must be determined from the signal itself. This is the problem of
clock recovery. A typical strategy samples several times per pulse and then uses
some criterion to pick the best one, to estimate the optimal time, or to interpolate an appropriate value. There must also be a way to deal with the situation
when the oscillator dening the symbol clock at the transmitter diers from the
oscillator dening the symbol clock at the receiver. Similarly, carrier synchronization is the process of recovering the carrier (in both frequency and phase) from
the received signal. This is the same task in a digital receiver as in an analog

34

Chapter 2. A Telecommunication System

design (recall that the cosine wave used to demodulate the received signal in
(2.6) was aligned in both phase and frequency with the modulating sinusoid at
the transmitter), though the details of implementation may dier.
In many applications (such as cell phones), messages come in clusters called
packets, and each packet has a header (that is located in some agreed-upon
place within each data block) that contains important information. The process
of identifying where the header appears in the received signal is called frame
synchronization, and is often implemented using a correlation technique.
The point of view adopted in Software Receiver Design is that many of
these synchronization tasks can be stated quite simply as optimization problems.
Accordingly, many of the standard solutions to synchronization tasks can be
viewed as solutions to these optimization problems.

r The problem of clock (or timing) recovery can be stated as that of nding

a timing oset to maximize the energy of the received signal. Solving this
optimization problem via a gradient technique leads to a standard algorithm
for timing recovery.
The problem of carrier phase synchronization can be stated as that of nding
a phase oset to minimize a particular function of the modulated received
signal. Solving this optimization problem via a gradient technique leads to
the phase-locked loop, a standard method of carrier recovery.
Carrier phase synchronization can also be stated using an alternative performance function that leads directly to the Costas loop, another standard
method of carrier recovery.

Our presentation focuses on solving problems using simple recursive (gradient)

methods. Once the synchronization problems are correctly stated, techniques
for their solution become obvious. With the exception of frame synchronization (which is approached via correlational methods) the problem of designing
synchronizers is unied via one simple concept, that of the minimization (or
maximization) of an appropriate performance function. Chapters 6, 10, and 12
contain details.

2.12

Equalization
When all is well in the digital receiver, there is no interaction between adjacent
data values and all frequencies are treated equally. In most real wireless systems
(and many wired systems as well), however, the transmission channel causes
multiple copies of the transmitted symbols, each scaled dierently, to arrive at
the receiver at dierent times. This intersymbol interference can garble the data.
The channel may also attenuate dierent frequencies by dierent amounts. Thus
frequency selectivity can render the data indecipherable.
The solution to both of these problems is to build a lter in the receiver that
attempts to undo the eects of the channel. This lter, called an equalizer, cannot

A Telecommunication System

35

be xed in advance by the system designer, however, because it must be dierent

in order to compensate for dierent channel paths that are encountered when the
system is operating. The problem of equalizer design can be stated as a simple
optimization problem, that of nding a set of lter parameters to minimize an
appropriate function of the error, given only the received data (and perhaps a
training sequence). This problem is investigated in detail in Chapter 13, where
the same kinds of adaptive techniques that are used to solve the synchronization
problems can also be applied to solve the equalization problem.

2.13

Decisions and Error Measures

In analog systems, the transmitted waveform can attain any value, but in a
digital implementation the transmitted message must be one of a small number
of values dened by the symbol alphabet. Consequently, the received waveform
in an analog system can attain any value, but in a digital implementation the
recovered message is meant to be one of a small number of values from the
source alphabet. Thus, when a signal is demodulated to a symbol and it is not
a member of the alphabet, the dierence between the demodulated value (called
a soft decision) and the nearest element of the alphabet (the hard decision) can
provide valuable information about the performance of the system.
To be concrete, label the signals at various points as shown in Figure 2.13.

r The binary input message b().

r The coded signal w() is a discrete-time sequence drawn from a nite alphabet.
r The signal m() at the output of the lter and equalizer is continuous-valued
at discrete times.

r Q{m()} is a version of m() that is quantized to the nearest member of the

alphabet.

r The decoded signal b() is the nal (binary) output of the receiver.
If all goes well and the message is transmitted, received, and decoded successfully, then the output should be the same as the input, although there may be
some delay between the time of transmission and the time when the output is
available. When the output diers from the message, then errors have occurred
during transmission.
There are several ways to measure the quality of the system. For instance, the
symbol recovery error
e(kT ) = w((k )T ) m(kT )
measures the dierence between the message and the soft decision. The average
squared error,
M
1  2
e (kT ),
M
k=1

36

Chapter 2. A Telecommunication System

Binary
message
sequence b

{3, 1, 1, 3}

Coding

P( f )
Pulse
shaping

Analog
upconversion
Carrier
specification

Antenna

Analog
Digital downPulseconversion
conversion
matched
to IF
to baseband
filter
Analog
Ts
received
Input to the
Carrier
signal
software synchronization
receiver

Other FDM
Noise
users

Transmitted
signal
Channel

Reconstructed
message
m Q(m) {3, 1, 1, 3} ^
T
b
Downsampling Equalizer Decision
Decoding
Timing
synchronization

Source and
error coding,
with frame
synchronization

Figure 2.13 PAM system diagram.

gives a measure of the performance of the system. This can be used as in Chapter
13 to adjust the parameters of an equalizer when the source message is known.
Alternatively, the dierence between the message w() and the quantized output
of the receiver Q{m()} can be used to measure the hard-decision error
e(kT ) = w((k )T ) Q{m(kT )}.
The decision-directed error replaces this with
e(kT ) = Q{m(kT )} m(kT ),
the error between the soft decisions and the associated hard decisions. This error
is used in Section 13.4 as a way to adjust the parameters in an equalizer when
the source message is unknown, as a way of adjusting the phase of the carrier in
Section 10.5, and as a way of adjusting the symbol timing in Section 12.3.
There are other useful indicators of the performance of digital communication
receivers. Let Tb be the bit duration (when there are two bits per symbol, Tb =
T /2). The indicator

1 if b((k )Tb ) = b(kTb )
c(kTb ) =
0 if b((k )Tb ) = b(kTb )
counts how many bits have been incorrectly received, and the bit error rate is
BER =

M
1 
c(kTb ).
M

(2.10)

k=1

Similarly, the symbol error rate sums the indicators


1 if w((k )T )) = Q{m(kT )}
c(kT ) =
0 if w((k )T )) = Q{m(kT )}
counting the number of alphabet symbols that were transmitted incorrectly.
More subjective or context-dependent measures are also possible, such as the

A Telecommunication System

37

percentage of typical listeners who can accurately decipher the output of the
receiver.
No matter what the exact form of the error measure, the ultimate goal is the
accurate and ecient transmission of the message.

2.14

Coding and Decoding

What is information? How much can move across a particular channel in a given
amount of time? Claude Shannon proposed a method of measuring information
in terms of bits, and a measure of the capacity of the channel in terms of the bit
ratethe number of bits transmitted per second. This is dened quantitatively
by the channel capacity, which is dependent on the bandwidth of the channel
and on the power of the noise in comparison with the power of the signal. For
most receivers, however, the reality is far from the capacity, and this is caused
by two factors. First, the data to be transmitted are often redundant, and the
redundancy squanders the capacity of the channel. Second, the noise can be
unevenly distributed among the symbols. When large noises disrupt the signal,
then excessive errors occur.
The problem of redundancy is addressed in Chapter 14 by source coding, which
strives to represent the data in the most concise manner possible. After demonstrating the redundancy and correlation of English text, Chapter 14 introduces
the Human code, which is a variable-length code that assigns short bit strings
to frequent symbols and longer bit strings to infrequent symbols. Like Morse
code, this will encode the letter e with a short code word, and the letter z
with a long code word. When correctly applied, the Human procedure can be
applied to any symbol set (not just the letters of the alphabet), and is nearly
optimal, that is, it approaches the limits set by Shannon.
The problem of reducing the sensitivity to noise is addressed in Chapter 14
using the idea of linear block codes, which cluster a number of symbols together,
and then add extra bits. A simple example is the (binary) parity check, which
adds an extra bit to each character. If there is an even number of ones then a 1
is added, and if there is an odd number of ones, a 0 is added. The receiver can
always detect that a single error has occurred by counting the number of ones
received. If the sum is even, then an error has occurred, while if the sum is odd
then no single error can have occurred. More sophisticated versions can not only
detect errors, but also correct them.
Like good equalization and proper synchronization, coding is an essential part
of the operation of digital receivers.

38

Chapter 2. A Telecommunication System

2.15

A Telecommunication System
The complete system diagram, including the digital receiver that will be built in
this text, is shown in Figure 2.13. This system includes the following.

r A source coding that reduces the redundancy of the message.

r An error coding that allows detection and/or correction of errors that may
r
r
r
r
r
r
r
r
r
r
r
r
r

occur during the transmission.

A message sequence of T -spaced symbols drawn from a nite alphabet.
Pulse shaping of the message, designed (in part) to conserve bandwidth.
Analog upconversion to the carrier frequency (within a specied tolerance).
Channel distortion of transmitted signal.
Summation with other FDM users, channel noise, and other interferers.
Analog downconversion to intermediate frequency (including bandpass preltering around the desired segment of the FDM passband).
A/D impulse sampling (preceded by antialiasing lter) at a rate of 1/Ts with
arbitrary start time. The sampling rate is assumed to be at least as fast as
the symbol rate 1/T .
Downconversion to baseband (requiring carrier-phase and frequency synchronization).
Lowpass (or pulse-shape-matched) ltering for the suppression of out-of-band
users and channel noise.
Downsampling with timing adjustment to T -spaced symbol estimates.
Equalization ltering to combat intersymbol interference and narrowband
interferers.
A decision device quantizing soft-decision outputs of the equalizer to the nearest member of the source alphabet (i.e., the hard decision).
Source and error decoders.

Of course, permutations and variations of this system are possible, but we

believe that Figure 2.13 captures the essence of many modern transmission systems.

2.16

Stairway to Radio
The path taken by Software Receiver Design is to break down the telecommunication system into its constituent elements: the modulators and demodulators,
the samplers and lters, the coders and decoders. In the various tasks within
each chapter, you are asked to build a simulation of the relevant piece of the system. In the early chapters, the parts need to operate only in a pristine, idealized
environment, but as we progress, impairments and noises inevitably intrude. The
design evolves to handle the increasingly realistic scenarios.

A Telecommunication System

39

Throughout this text, we ask you to consider a variety of small questions, some
of which are mathematical in nature, most of which are what if questions best
answered by trial and simulation. We hope that this combination of reection
and activity will be useful in enlarging your understanding and in training your
intuition.

For Further Reading

There are many books about various aspects of communication systems. Here
are some of our favorites. Three basic texts that utilize probability from the
outset, and that also pay substantial attention to pragmatic design issues (such
as synchronization) are the following:

r J. B. Anderson, Digital Transmission Engineering, IEEE Press, 1999;

r J. G. Proakis and M. Salehi, Digital Communications, 5th edition, McGraw-

Hill, 2007. [This text also has a Matlab-based companion, J. G. Proakis, M.

Salehi, and G. Bauch, Contemporary Communication Systems Using Matlab,
2nd edition, Cengage Learning, 2004];
S. Haykin, Communication Systems, 4th edition, John Wiley and Sons, 2001.

Three introductory texts that delay the introduction of probability until the
latter chapters are the following:

r L. W. Couch III, Digital and Analog Communication Systems, 6th edition,

Prentice-Hall, 2001;

r B. P. Lathi, Modern Digital and Analog Communication Systems, 3rd edition,

Oxford University Press, 1998;

r F. G. Stremler, Introduction to Communication Systems, 3rd edition, Addison

Wesley, 1990.
These nal three references are probably the most compatible with Software
Receiver Design in terms of the assumed mathematical background.

The Six Elements

At rst glance, block diagrams such as the communication system shown in Figure 2.13 probably appear complex and intimidating. There are so many dierent
blocks and so many unfamiliar names and acronyms! Fortunately, all the blocks
can be built from six simple elements:

r signal generators such as oscillators, which create sine and cosine waves,
r linear time-invariant lters, which augment or diminish the amplitude of particular frequencies or frequency ranges in a signal,

r samplers, which change analog (continuous-time) signals into discrete-time

signals,

r static nonlinearities such as squarers and quantizers, which can add frequency
content to a signal,

r linear time-varying systems such as mixers that shift frequencies around in

useful and understandable ways, and

r adaptive elements, which track the desired values of parameters as they slowly
change over time.
This section provides a brief overview of these six elements. In doing so, it also
reviews some of the key ideas from signals and systems. Later chapters explore
how the elements work, how they can be modied to accomplish particular tasks
within the communication system, and how they can be combined to create a
large variety of blocks such as those that appear in Figure 2.13.
The elements of a communication system have inputs and outputs; the element
itself operates on its input signal to create its output signal. The signals that
form the inputs and outputs are functions that represent the dependence of some
variable of interest (such as a voltage, current, power, air pressure, temperature,
etc.) on time.
The action of an element can be described by the manner in which it operates
in the time domain, that is, how the element changes the input waveform
moment by moment into the output waveform. Another way of describing the
action of an element is by how it operates in the frequency domain, that is,

41

Amplitude

Amplitude

The Six Elements

Time
Input signal as a
function of time
x(t)
X( f )

Time
Output signal as a
function of time
y(t)
Y( f )

Element

Frequency
The magnitude spectrum
shows the frequency context
of the input signal

Output signal as a
function of frequency
Magnitude

Magnitude

Input signal as a
function of frequency

Figure 3.1 The element

transforms the input signal x

into the output signal y. The
action of an element can be
thought of in terms of its
eect on the signals in time,
or (via the Fourier transform)
in terms of its eect on the
spectra of the signals.

Frequency
The magnitude spectrum
shows the frequency content
of the output signal

by how the frequency content of the input relates to the frequency content of
the output. Figure 3.1 illustrates these two complementary ways of viewing the
elements. Understanding both the time-domain and frequency-domain behavior
is essential. Accordingly, the following sections describe the action of the six
elements in both time and frequency.
Readers who have studied signals and systems (as is often required in electrical engineering degrees) will recognize that the time-domain representation
of a signal and its frequency-domain representation are related by the Fourier
transform, which is briey reviewed in the next section.

3.1

Finding the Spectrum of a Signal

A signal s(t) can often be expressed in analytical form as a function of time t,
and the Fourier transform is dened as in (2.1) as the integral of s(t)e2jf t . The
resulting transform S(f ) is a function of frequency. S(f ) is called the spectrum of
the signal s(t) and describes the frequencies present in the signal. For example,
if the time signal is created as a sum of three sine waves, the spectrum will
have spikes corresponding to each of the constituent sines. If the time signal
contains only frequencies between 100 and 200 Hz, the spectrum will be zero for
all frequencies outside of this range. A brief guide to Fourier transforms appears
in Appendix D, and a summary of all the transforms and properties that are
used throughout Software Receiver Design appears in Appendix A.
Often, however, there is no analytical expression for a signal; that is, there is
no (known) equation that represents the value of the signal over time. Instead,

42

Chapter 3. The Six Elements

the signal is dened by measurements of some physical process. For instance, the
signal might be the waveform at the input to the receiver, the output of a linear
lter, or a sound waveform encoded as an MP3 le. In all these cases, it is not
possible to nd the spectrum by calculating a Fourier transform since the signal
is not known in analytical form.
Rather, the discrete Fourier transform (and its cousin, the more rapidly computable fast Fourier transform, or FFT) can be used to nd the spectrum or frequency content of a measured signal. The Matlab function plotspec.m, which
plots the spectrum of a signal, is available on the website. Its help le1 notes
% p l o t s p e c ( x , Ts ) p l o t s th e spectrum o f x
% Ts=time ( i n s e c s ) between a d j a c e n t s a m p l e s

The function plotspec.m is easy to use. For instance, the spectrum of a square
wave can be found using the following sequence:
Listing 3.1. specsquare.m plot the spectrum of a square wave
f =10;
time =2;
Ts =1/1000;
t=Ts : Ts : time ;
x=sign ( cos (2 * pi * f * t ) ) ;
p l o t s p e c ( x , Ts )

%
%
%
%
%
%

frequency of square
l e n g t h o f time
time i n t e r v a l between
c r e a t e a time v e c t o r
s q u a r e wave = s i g n o f
c a l l p l o t s p e c to draw

wave
s a m p l es
c o s wave
spectrum

The output of specsquare.m is shown2 in Figure 3.2. The top plot shows time=2
seconds of a square wave with f=10 cycles per second. The bottom plot shows a
series of spikes that dene the frequency content. In this case, the largest spike
occurs at 10 Hz, followed by smaller spikes at all the odd-integer multiples (i.e.,
at 30, 50, 70, etc.).
Similarly, the spectrum of a noise signal can be calculated as
Listing 3.2. specnoise.m plot the spectrum of a noise signal
time =1;
Ts =1/10000;
x=randn ( 1 , time /Ts ) ;
p l o t s p e c ( x , Ts )

%
%
%
%

l e n g t h o f time
time i n t e r v a l between s a m p l es
Ts p o i n t s o f n o i s e f o r time s e c o n d s
c a l l p l o t s p e c to draw spectrum

A typical run of specnoise.m is shown in Figure 3.3. The top plot shows
the noisy signal as a function of time, while the bottom shows the magnitude
spectrum. Because successive values of the noise are generated independently, all
frequencies are roughly equal in magnitude. Each run of specnoise.m produces
plots that are qualitatively similar, though the details will dier.
1

You can view the help le for the Matlab function xxx by typing help xxx at the Matlab
prompt. If you get an error such as xxx not found, then this means either that the function
does not exist, or that it needs to be moved into the same directory as the Matlab application. If you dont know what the proper command to do a job is, then use lookfor. For
instance, to nd the command that inverts a matrix, type lookfor inverse. You will nd
the desired command inv.
All code listings in Software Receiver Design can be found on the website. We encourage
you to open Matlab and explore the code as you read.

Amplitude

The Six Elements

1
0.5
0
0.5
1
0

Magnitude

43

0.2 0.4

0.6

0.8 1 1.2 1.4

Seconds

1.6

1.8

1200

Figure 3.2 A square wave and

its spectrum, as calculated by

using plotspec.m.

800
400
0

400 300 200 100 0 100 200 300 400

Frequency

Exercise 3.1. Use specsquare.m to investigate the relationship between the

time behavior of the square wave and its spectrum. The Matlab command
zoom on is often helpful for viewing details of the plots.
a. Try square waves with dierent frequencies: f=20, 40, 100, 300 Hz. How
do the time plots change? How do the spectra change?
b. Try square waves of dierent lengths, time=1, 10, 100 seconds. How does
the spectrum change in each case?
c. Try dierent sampling times, Ts=1/100, 1/10000 seconds. How does the
spectrum change in each case?
Exercise 3.2. In your signals and systems course, you probably calculated (analytically) the spectrum of a square wave by using the Fourier series. How does
this calculation compare with the discrete data version found by specsquare.m?
Exercise 3.3. Mimic the code in specsquare.m to nd the spectrum of
a.
b.
c.
d.
e.

an exponential pulse s(t) = et for 0 < t < 10,

a scaled exponential pulse s(t) = 5et for 0 < t < 10,
2
a Gaussian pulse s(t) = et for 2 < t < 2,
2
a Gaussian pulse s(t) = et for 20 < t < 20,
the sinusoids s(t) = sin(2f t + ) for f = 20, 100, 1000, with = 0, /4, /2,
and 0 < t < 10.

Exercise 3.4. Matlab has several commands that create random numbers.
a. Use rand to create a signal that is uniformly distributed on [1, 1]. Find the
spectrum of the signal by mimicking the code in specnoise.m.
b. Use rand and the sign function to create a signal that is +1 with probability
1/2 and 1 with probability 1/2. Find the spectrum of the signal.
c. Use randn to create a signal that is normally distributed with mean 0 and
variance 3. Find the spectrum of the signal.

44

Chapter 3. The Six Elements

Amplitude

4
2
0
2

Magnitude

0.2

0.4
0.6
Seconds

0.8

Figure 3.3 A noise signal and its

spectrum, as calculated using

plotspec.m.

300
200
100
0

4000

2000

0
Frequency

2000

4000

Exercise 3.5. Modify the code in plotspec.m to also plot the phase spectrum.
a. Plot the phase spectrum of a sine wave and a cosine wave, both of the same
frequency. How do they dier?
b. Plot the phase spectrum of the random signal created in Exercise 3.4(a).
While plotspec.m can be quite useful, ultimately, it will be necessary to have
more exibility, which, in turn, requires one to understand how the FFT function
inside plotspec.m works. This will be discussed at length in Chapter 7. The
next six sections describe the six elements that are at the heart of communications systems. The elements are described both in the time domain and in the
frequency domain.

3.2

The First Element: Oscillators

The Latin word oscillare means to ride in a swing. It is the origin of oscillate,
which means to move back and forth in steady unvarying rhythm. Thus, a device
that creates a signal that moves back and forth in a steady, unvarying rhythm is
called an oscillator. An electronic oscillator is a device that produces a repetitive
electronic signal, usually a sinusoidal wave.
A basic oscillator is diagrammed in Figure 3.4. Oscillators are typically
designed to operate at a specied frequency f0 , and the input species the phase
(t) of the output
s(t) = cos(2f0 t + (t)).
The input may be a xed number, but it may also be a signal; that is, it may
change over time. In this case, the output is no longer a pure sinusoid of frequency
f0 . For instance, suppose the phase is a ramp or line with slope 2c; that is,
(t) = 2ct. Then s(t) = cos(2f0 t + 2ct) = cos(2(f0 + c)t), and the actual
frequency of oscillation is f0 + c.

The Six Elements

cos(2f0t+(t))

(t)

Figure 3.4 An oscillator creates a sinusoidal oscillation

with a specied frequency f0 and input (t).

Amplitude

1
0.5
0
0.5
1
120

Magnitude

45

Figure 3.5 A sinusoidal oscillator

0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8

Seconds

80

creates a signal that can be

viewed in the time domain as in
the top plot, or in the frequency
domain as in the bottom plot.

40
0
50 40 30 20 10 0 10
Frequency

20

30

40

50

There are many ways to build oscillators from analog components. Generally,
there is an amplier and a feedback circuit that returns a portion of the amplied
wave back to the input. When the feedback is aligned properly in phase, sustained
oscillations occur.
Digital oscillators are simpler. Since they can be directly calculated, no amplier or feedback is needed. For example, a digital sine wave of frequency f Hz
and a phase of radians can be represented mathematically as
s(kTs ) = cos(2f kTs + ),

(3.1)

where Ts is the time between samples and where k is an integer counter k =

1, 2, 3, . . .. Equation (3.1) can be directly implemented in Matlab:
Listing 3.3. speccos.m plot the spectrum of a cosine wave
f =10; p h i =0;
time =2;
Ts =1/100;
t=Ts : Ts : time ;
x=cos (2 * pi * f * t+p h i ) ;
p l o t s p e c ( x , Ts )

%
%
%
%
%
%

s p e c i f y f r e q u e n c y and phase
l e n g t h o f time
time i n t e r v a l between s a m p l es
c r e a t e a time v e c t o r
c r e a t e c o s wave
draw waveform and spectrum

The output of speccos.m is shown in Figure 3.5. As expected, the time plot
shows an undulating sinusoidal signal with f = 10 repetitions in each second.
The spectrum shows two spikes, one at f = 10 Hz and one at f = 10 Hz. Why
are there two spikes? Basic Fourier theory shows that the Fourier transform of
a cosine wave is a pair of delta functions at plus and minus the frequency of the
cosine wave (see Equation (A.18)). The two spikes of Figure 3.5 mirror these
two delta functions. Alternatively, recall that a cosine wave can be written using
Eulers formula as the sum of two complex exponentials, as in (A.2). The spikes of
Figure 3.5 represent the magnitudes of these two (complex-valued) exponentials.

46

Chapter 3. The Six Elements

Exercise 3.6. Mimic the code in speccos.m to nd the spectrum of a cosine

wave
a. for dierent frequencies f=1, 2, 20, 30 Hz,
b. for dierent phases = 0, 0.1, /8, /2 radians,
c. for dierent sampling rates Ts=1/10, 1/1000, 1/100000.
Exercise 3.7. Let x1 (t) be a cosine wave of frequency f = 10, x2 (t) be a cosine
wave of frequency f = 18, and x3 (t) be a cosine wave of frequency f = 33. Let
x(t) = x1 (t) + 0.5 x2 (t) + 2 x3 (t). Find the spectrum of x(t). What property
of the Fourier transform does this illustrate?
Exercise 3.8. Find the spectrum of a cosine wave in the following cases.
a.
b.
c.
d.

3.3

f
f

is
is
is
is

a
a
a
a

function
function
function
function

of
of
of
of

time.
time.
time.
time.

Try
Try
Try
Try

(t) = 10t.
(t) = t2 .
f (t) = sin(2t).
f (t) = t2 .

The Second Element: Linear Filters

Linear time-invariant lters shape the spectrum of a signal. If the signal has
too much energy in the low frequencies, a highpass lter can remove them. If
the signal has too much high-frequency noise, a lowpass lter can reject it. If a
signal of interest resides only between f and f , then a bandpass lter tuned
to pass frequencies between f and f can remove out-of-band interference and
noise. More generally, suppose that a signal has frequency bands in which the
magnitude of the spectrum is lower than desired and other bands in which the
magnitude is greater than desired. Then a linear lter can compensate by increasing or decreasing the magnitude. This section provides an overview of how to
implement simple lters in Matlab. More thorough treatments of the theory,
design, use, and implementation of lters are given in Chapter 7.
While the calculations of a linear lter are usually carried out in the time
domain, lters are often specied in the frequency domain. Indeed, the words
used to specify lters (such as lowpass, highpass, and bandpass) describe how the
lter acts on the frequency content of its input. Figure 3.6, for instance, shows
a noisy input entering three dierent lters. The frequency response of the LPF
shows that it allows low frequencies (those below the cuto frequency f ) to pass,
while removing all frequencies above the cuto. Similarly, the HPF passes all the
high frequencies and rejects those below its cuto f . The frequency response of
the BPF is specied by two frequencies. It will remove all frequencies below f
and all frequencies above f , leaving only the region between.

The Six Elements

47

X( f ) Mag.
Figure 3.6 A white

Y1( f )
Freq.
White input

f*

f*

y1(t)

LPF

f*

x(t)

f*

Y2( f )
f*f*

f* f*

BPF

y2(t)

f*f*

f* f*

Y3( f )
f*

f*

f*
HPF

f*

y3(t)

signal containing all

frequencies is passed
through a lowpass lter
(LPF) leaving only the
low frequencies, a
bandpass lter (BPF)
leaving only the middle
frequencies, and a
highpass lter (HPF)
leaving only the high
frequencies.

Figure 3.6 shows the action of ideal lters. How close are actual implementations? The Matlab code in filternoise.m shows that it is possible to create
digital lters that reliably and accurately carry out these tasks.
Listing 3.4. filternoise.m lter a noisy signal three ways
time =3;
Ts =1/10000;
x=randn ( 1 , time /Ts ) ;
f i g u r e ( 1 ) , p l o t s p e c ( x , Ts )
f r e q s =[0 0 . 2 0 . 2 1 1 ] ;
amps=[1 1 0 0 ] ;
b=f i r p m ( 1 0 0 , f r e q s , amps ) ;
y l p=f i l t e r ( b , 1 , x ) ;
f i g u r e ( 2 ) , p l o t s p e c ( ylp , Ts )
f r e q s =[0 0 . 2 4 0 . 2 6 0 . 5 0 . 5 1 1 ] ;
amps=[0 0 1 1 0 0 ] ;
b=f i r p m ( 1 0 0 , f r e q s , amps ) ;
ybp=f i l t e r ( b , 1 , x ) ;
f i g u r e ( 3 ) , p l o t s p e c ( ybp , Ts )
f r e q s =[0 0 . 7 4 0 . 7 6 1 ] ;
amps=[0 0 1 1 ] ;
b=f i r p m ( 1 0 0 , f r e q s , amps ) ;
yhp=f i l t e r ( b , 1 , x ) ;
f i g u r e ( 4 ) , p l o t s p e c ( yhp , Ts )

%
%
%
%

l e n g t h o f time
time i n t e r v a l between s a m p l es
generate noise s i gnal
draw spectrum o f i n p u t

% s p e c i f y th e LP f i l t e r
% do th e f i l t e r i n g
% p l o t th e o u t p u t spectr um

% BP f i l t e r
% do th e f i l t e r i n g
% p l o t th e o u t p u t spectr um

% s p e c i f y th e HP f i l t e r
% do th e f i l t e r i n g
% p l o t th e o u t p u t spectr um

The output of filternoise.m is shown in Figure 3.7. Observe that the spectra at the output of the lters are close approximations to the ideals shown in
Figure 3.6. There are some dierences, however. While the idealized spectra are
completely at in the passband, the actual ones are rippled. While the idealized
spectra completely reject the out-of-band frequencies, the actual ones have small
(but nonzero) energy at all frequencies. Two new Matlab commands are used
in filternoise.m. The firpm command species the contour of the lter as a
line graph (Octave and early versions of Matlab call this command remez). For
instance, typing
plot ( [ 0 0 . 2 4 0 . 2 6 0 . 5 0 . 5 1 1 ] , [ 0 0 1 1 0 0 ] )

48

Chapter 3. The Six Elements

Magnitude spectrum at input

Magnitude spectrum at output of lowpass filter

Magnitude spectrum at output of bandpass filter

Figure 3.7 The spectrum of a white

signal containing all frequencies is

shown in the top gure. This is
passed through three lters: a
lowpass, a bandpass, and a highpass.
The spectra at the outputs of these
three lters are shown in the second,
third, and bottom plots. The actual
lters behave much like their
idealized counterparts in Figure 3.6.

4000
2000
0
2000
4000
Magnitude spectrum at output of highpass filter

at the Matlab prompt draws a box that represents the action of the BPF
designed in filternoise.m (over the positive frequencies). The frequencies
are specied as percentages of fN Y Q = 1/(2Ts ), which in this case is equal to
5000 Hz. (fN Y Q is discussed further in the next section.) Thus the BPF in
filternoise.m passes frequencies between 0.265000 Hz and 0.55000 Hz, and
rejects all others. The filter command uses the output of firpm to carry out
the ltering operation on the vector specied in its third argument. More details
about these commands are given in the section on practical ltering in Chapter 7.
Exercise 3.9. Mimic the code in filternoise.m to create a lter that
a. passes all frequencies above 500 Hz,
b. passes all frequencies below 3000 Hz,
c. rejects all frequencies between 1500 and 2500 Hz.
Exercise 3.10. Change the sampling rate to Ts=1/20000. Redesign the three
lters from Exercise 3.9.
Exercise 3.11. Let x1 (t) be a cosine wave of frequency f = 800, x2 (t) be a cosine
wave of frequency f = 2000, and x3 (t) be a cosine wave of frequency f = 4500.
Let x(t) = x1 (t) + 0.5 x2 (t) + 2 x3 (t). Use x(t) as input to each of the three
lters in filternoise.m. Plot the spectra, and explain what you see.
Exercise 3.12. TRUE or FALSE: A linear, time-invariant system exists that
has input a cos(bt) and output c sin(dt) with a = c and |b| = |d|. Explain.
Exercise 3.13. TRUE or FALSE: Filtering a passband signal with absolute
bandwidth B through certain xed linear lters can result in an absolute bandwidth of the lter output greater than B. Explain.

49

The Six Elements

Exercise 3.14. TRUE or FALSE: A linear, time-invariant, nite-impulseresponse lter with a frequency response having unit magnitude over all frequencies and a straight-line, sloped phase curve has as its transfer function a
pure delay. Explain.
Exercise 3.15. TRUE or FALSE: Processing a bandlimited signal through a
linear, time-invariant lter can increase its half-power bandwidth. Explain.

3.4

The Third Element: Samplers

Since part of any digital transmission system is analog (transmissions through
the air, across a cable, or along a wire are inherently analog), and part of the
system is digital, there must be a way to translate the continuous-time signal
into a discrete-time signal and vice versa. The process of sampling an analog signal, sometimes called analog-to-digital conversion, is easy to visualize in the time
domain. Figure 3.8 shows how sampling can be viewed as the process of evaluating a continuous-time signal at a sequence of uniformly spaced time intervals,
thus transforming the analog signal x(t) into the discrete-time signal x(kTs ).
One of the key ideas in signals and systems is the Fourier series: a signal is
periodic in time (it repeats every P seconds) if and only if the spectrum can
be written as a sum of complex sinusoids with frequencies at integer multiples
of a fundamental frequency f . Moreover, this fundamental frequency can be
written in terms of the period as f = 1/P . Thus, if a signal repeats 100 times
every second (P = 0.01 s), then its spectrum consists of a sum of sinusoids with
frequencies 100, 200, 300, . . . Hz.
Conversely, if a spectrum is built from a sum of sinusoids with frequencies
100, 200, 300, . . . Hz, then it must represent a periodic signal in time that has
period P = 0.01 s. Said another way, the nonzero portions of the spectrum are
uniformly spaced f = 100 Hz apart. This uniform spacing can be interpreted as
a sampling (in frequency) of an underlying continuous-valued spectrum. This is

=
T s)

]
x[1

x(

)=

]
x[2

Ts

x(2

x(t)

3
x[

Ts

]
x[4

Ts

x(4

)=
Ts

x(kTs) = x[k]

3
x(

x(t)

)=

(a)

Ts

2Ts

3Ts

4Ts

(b)
Figure 3.8 The sampling process is shown in (b) as an evaluation of the signal x(t) at
times . . . , 2Ts , Ts , 0, Ts , 2Ts , . . .. This procedure is schematized in (a) as an element
that has the continuous-time signal x(t) as input and the discrete-time signal x(kTs )
as output.

50

Chapter 3. The Six Elements

Uniform sampling in frequency

f

Periodic in time
P
f=
...

...

1
P
f 2f 3f 4f 5f

Fourier Transform

Ts

Inverse Fourier Transform

fs

...
fs =
Uniform sampling in time

1 3f
s
Ts
2

...
fs
2

fs
2

3fs
2

Periodic in frequency

Figure 3.9 Fouriers result says that any signal that is periodic in time has a spectrum

that consists of a collection of spikes uniformly spaced in frequency. Analogously, any

signal whose spectrum is periodic in frequency can be represented in time as a
collection of spikes uniformly spaced in time, and vice versa.

illustrated in the top portion of Figure 3.9, which shows the time-domain representation on the left and the corresponding frequency-domain representation on
the right.
The basic insight from Fourier series is that any signal which is periodic in
time can be reexpressed as a collection of uniformly spaced spikes in frequency;
that is,
Periodic in Time Uniform Sampling in Frequency.
The same arguments show the basic result of sampling, which is that
Uniform Sampling in Time Periodic in Frequency.
Thus, whenever a signal is uniformly sampled in time (say, with sampling interval
Ts seconds), the spectrum will be periodic; that is, it will repeat every fs = 1/Ts
Hz.
Two conventions are often observed when drawing periodic spectra that arise
from sampling. First, the spectrum is usually drawn centered at 0 Hz. Thus, if
the period of repetition is fs , this is drawn from fs /2 to fs /2, rather than from
0 to fs . This makes sense because the spectra of individual real-valued sinusoidal
components contain two spikes symmetrically located around 0 Hz (as we saw in
Section 3.2). Accordingly, the highest frequency that can be represented unambiguously is fs /2, and this frequency is often called the Nyquist frequency fN Y Q .
The second convention is to draw only one period of the spectrum. After all,
the others are identical copies that contain no new information. This is evident

The Six Elements

x[k]

y[i]

51

Figure 3.10 The discrete signal x[k] is downsampled by a factor

of m by removing all but one of every m samples. The

resulting signal is y[i], which takes on values y[i] = x[k]
whenever k = im + n.

in the bottom right diagram of Figure 3.9 where the spectrum between 3fs /2
and fs /2 is the same as the spectrum between fs /2 and 3fs /2. In fact, we
have been observing this convention throughout Sections 3.2 and 3.3, since all
of the gures of spectra (Figures 3.2, 3.3, 3.5, and 3.7) show just one period of
the complete spectrum.
Perhaps you noticed that plotspec.m changes the frequency axis when the
sampling interval Ts is changed. (If not, go back and redo Exercise 3.1(c).) By
the second convention, plotspec.m shows exactly one period of the complete
spectrum. By the rst convention, the plots are labeled from fN Y Q to fN Y Q .
What happens when the frequency of the signal is too high for the sampling
rate? The representation becomes ambiguous. This is called aliasing, and is investigated by simulation in the problems below. Aliasing and other sampling-related
issues (such as reconstructing an analog signal from its samples) are covered in
more depth in Chapter 6.
Closely related to the digital sampling of an analog signal is the (digital)
downsampling of a digital signal, which changes the rate at which the signal is
represented. The simplest case downsamples by a factor of m, removing all but
one out of every m samples. This can be written
y[i] = x[im + n],
where n is an integer between 0 and m 1. For example, with m = 3 and n = 1,
y[i] is the sequence that consists of every third value of x[k],
y[0] = x[1], y[1] = x[4], y[2] = x[7], y[3] = x[10], etc.
This is commonly drawn in block form as in Figure 3.10. If the spectrum of x[k]
is bandlimited to 1/m of the Nyquist rate, then downsampling by m loses no
information. Otherwise, aliasing occurs. Like analog-to-digital sampling, downsampling is a time-varying operation.
Exercise 3.16. Mimicking the code in speccos.m with the sampling interval
Ts=1/100, nd the spectrum of a cosine wave cos(2f t) when f=30, 40, 49,
50, 51, 60 Hz. Which of these show aliasing?
Exercise 3.17. Create a cosine wave with frequency 50 Hz. Plot the spectrum when this wave is sampled at Ts=1/50, 1/90, 1/100, 1/110, and 1/200.
Which of these show aliasing?
Exercise 3.18. Mimic the code in speccos.m with sampling interval Ts=1/100
to nd the spectrum of a square wave with fundamental f=10, 20, 30, 33,

52

Chapter 3. The Six Elements

43 Hz. Can you predict where the spikes will occur in each case? Which of the
square waves show aliasing?

3.5

The Fourth Element: Static Nonlinearities

Linear systems3 such as lters cannot add new frequencies to a signal, though
they can remove unwanted frequencies. Nonlinearities such as squaring and quantizing can and will add new frequencies. These can be useful in the communication system in a variety of ways.
Perhaps the simplest (memoryless) nonlinearity is the square, which takes
the input at each time instant and multiplies it by itself. Suppose the input
is a sinusoid at frequency f , that is, x(t) = cos(2f t). Then the output is the
sinusoid squared, which can be rewritten using the cosinecosine product (A.4)
as
1 1
y(t) = x2 (t) = cos2 (2f t) = + cos(2(2f )t).
2 2
The spectrum of y(t) has a spike at 0 Hz due to the constant, and a spike at
2f Hz from the double-frequency term. Unfortunately, the action of a squaring
element is not always as simple as this example might suggest. The following
exercises encourage you to explore the kinds of changes that occur in the spectra
when using a variety of simple nonlinear elements.
Exercise 3.19. Mimic the code in speccos.m with Ts=1/1000 to nd the
spectrum of the output y(t) of a squaring block when the input is
a. x(t) = cos(2f t) for f = 100 Hz,
b. x(t) = cos(2f1 t) + cos(2f2 t) for f1 = 100 and f2 = 150 Hz,
c. a ltered noise sequence with nonzero spectrum between f1 = 100 and f2 =
300 Hz. Hint: generate the input by modifying filternoise.m.
d. Can you explain the large DC (zero-frequency) component?
Exercise 3.20. TRUE or FALSE: The bandwidth of x4 (t) cannot be greater
than that of x(t). Explain.
Exercise 3.21. Try dierent values of f1 and f2 in Exercise 3.19. Can you
predict what frequencies will occur in the output? When is aliasing an issue?
Exercise 3.22. Repeat Exercise 3.21 when the input is a sum of three sinusoids.
3

To be accurate, these systems must be exponentially stable and time-invariant.

The Six Elements

53

Exercise 3.23. Suppose that the output of a nonlinear block is the rectication
(absolute value) of the input y(t) = |x(t)|. Find the spectrum of the output when
the input is
a. x(t) = cos(2f t) for f = 100 Hz,
b. x(t) = cos(2f1 t) + cos(2f2 t) for f1 = 100 and f2 = 125 Hz.
c. Repeat (b) for f1 = 110 and f2 = 200 Hz. Can you predict what frequencies
will be present for any f1 and f2 ?
d. What frequencies will be present if x(t) is the sum of three sinusoids f1 , f2 ,
and f3 ?
Exercise 3.24. Suppose that the output of a nonlinear block is y(t) = g(x(t)),
where

1 x(t) > 0
g(t) =
1 x(t) 0
is a quantizer that outputs plus one when the input is positive and minus one
when the input is negative. Find the spectrum of the output when the input is
a. x(t) = cos(2f t) for f = 100 Hz,
b. x(t) = cos(2f1 t) + cos(2f2 t) for f1 = 100 and f2 = 150 Hz.
Exercise 3.25. Quantization of an input is another kind of common nonlinearity. The Matlab function quantalph.m (available on the website) quantizes a
signal to the nearest element of a desired set. Its help le reads
%
%
%
%
%
%
%
%
%

y=q u a n t a l p h ( x , a l p h a b e t )
q u a n t i z e th e i n p u t s i g n a l x to th e a l p h a b e t
u s i n g n e a r e s t n e i g h b o r method
i n p u t x v e c t o r to be q u a n t i z e d
alphabet vector of d i s c r e t e values
t h a t y can assume
sorted in ascending order
output y quantized v ecto r

Let x be a random vector x=randn(1,n) of length n. Quantize x to the nearest

[3, 1, 1, 3].
a. What percentage of the outputs are 1s? 3s?
b. Plot the magnitude spectrum of x and the magnitude spectrum of the output.
c. Now let x=3*randn(1,n) and answer the same questions.

3.6

The Fifth Element: Mixers

One feature of most telecommunications systems is the ability to change the
frequency of the signal without changing its information content. For example,

54

Chapter 3. The Six Elements

speech occurs in the range below about 8 kHz. In order to transmit this, it is
upconverted (as in Section 2.3) to radio frequencies at which the energy can
easily propagate over long distances. At the receiver, it is downconverted (as in
Section 2.6) to the original frequencies. Thus the spectrum is shifted twice.
One way of accomplishing this kind of frequency shifting is to multiply the
signal by a cosine wave, as shown in Figure 3.11. The following Matlab code
implements a simple modulation.
Listing 3.5. modulate.m change the frequency of the input
time = . 5 ; Ts =1/10000;
t=Ts : Ts : time ;
f c =1000; cmod=cos (2 * pi * f c * t ) ;
f i =100; x=cos (2 * pi * f i * t ) ;
y=cmod . * x ;
f i g u r e ( 1 ) , p l o t s p e c ( cmod , Ts )
f i g u r e ( 2 ) , p l o t s p e c ( x , Ts )
f i g u r e ( 3 ) , p l o t s p e c ( y , Ts )

%
%
%
%
%
%

time and s a m p l i n g i n t e r v a l
d e f i n e a time v e c t o r
cr eate cos of f r eq f c
input i s cos of f r eq f i
m u l t i p l y i n p u t by cmod
f i n d s p e c t r a and p l o t

The rst three lines of the code create the modulating sinusoid (i.e., an oscillator).
The next line species the input (in this case another cosine wave). The Matlab
syntax .* calculates a point-by-point multiplication of the two vectors cmod and
x.
The output of modulate.m is shown in Figure 3.12. The spectrum of the input
contains spikes representing the input sinusoid at 100 Hz and the spectrum
of the modulating sinusoid contains spikes at 1000 Hz. As expected from the
modulation property of the transform, the output contains sinusoids at 1000
100 Hz, which appear in the spectrum as the two pairs of spikes at 900 and
1100 Hz. Of course, this modulation can be applied to any signal, not just to
an input sinusoid. In all cases, the output will contain two copies of the input,
one shifted up in frequency and the other shifted down in frequency.
Exercise 3.26. Mimic the code in modulate.m to nd the spectrum of the output
y(t) of a modulator block (with modulation frequency fc = 1000 Hz) when
a.
b.
c.
d.
e.

the
the
the
the
the

input
input
input
input
input

is
is
is
is
is

x(t) = cos(2f1 t) + cos(2f2 t) for f1 = 100 and f2 = 150 Hz,

a square wave with fundamental f = 150 Hz,
a noise signal with all energy below 300 Hz,
a noise signal bandlimited to between 2000 and 2300 Hz,
a noise signal with all energy below 1500 Hz.

x(t)

cos(2f0t+)

Figure 3.11 The mixing operation shifts all

x(t) cos(2f0t+)

frequencies of a signal x(t) by an amount

dened by the frequency f0 of the
modulating sinusoidal wave.

The Six Elements

55

Magnitude spectrum at input

Magnitude spectrum of the oscillator

3000 2000 1000 0

1000 2000 3000
Magnitude spectrum at output
Figure 3.12 The spectrum of the input sinusoid is shown in the top gure. The middle
gure shows the spectrum of the modulating wave. The bottom shows the spectrum
of the point-by-point multiplication (in time) of the two, which is the same as their
convolution (in frequency).

3.7

The Sixth Element: Adaptation

Adaptation is a primitive form of learning. Adaptive elements in telecommunication systems nd approximate values for unknown parameters in an attempt
to compensate for changing conditions or to improve performance. A common
strategy in parameter-estimation problems is to guess a value, assess how good
the guess is, and rene the guess over time. With luck, the guesses converge to
a useful estimate of the unknown value.
Figure 3.13 shows an adaptive element containing two parts. The adaptive
subsystem parameterized by a changes the input into the output. The qualityassessment mechanism monitors the output (and other relevant signals) and tries
to determine whether a should be increased or decreased. The arrow through the
system indicates that the a value is then adjusted accordingly.
Adaptive elements occur in a number of places in the communication system,
including the following.

r In an automatic gain control, the adaptive subsystem is multiplication by a

constant a. The quality-assessment mechanism gauges whether the power at
the output of the AGC is too large or too small, and adjusts a accordingly.

Input

Output

Quality
Assessment

Figure 3.13 The adaptive element is a

subsystem that transforms the input
into the output (parameterized by a)
and a quality-assessment mechanism
that evaluates how to alter a; in this
case, whether to increase or decrease
a.

56

Chapter 3. The Six Elements

r In a phase-locked loop, the adaptive subsystem contains a sinusoid with

r
r

an unknown phase shift a. The quality-assessment mechanism adjusts a to

maximize a ltered version of the product of the sinusoid and its input.
In a timing recovery setting, the adaptive subsystem is a fractional delay
given by a. One mechanism for assessing quality monitors the power of the
output, and adjusts a to maximize this power.
In an equalizer, the adaptive subsystem is a linear lter parameterized by
a set of as. The quality-assessment mechanism monitors the deviation of the
output of the system from a target set and adapts the as accordingly.

Chapter 6 provides an introduction to adaptive elements in communication

systems, and a detailed discussion of their implementation is postponed until
then.

3.8

Summary
The bewildering array of blocks and acronyms in a typical communication system
diagram really consists of just a handful4 of simple elements: oscillators, linear
lters, samplers, static nonlinearities, mixers, and adaptive elements. For the
most part, these are ideas that the reader will have encountered to some degree in
previous studies, but they have been summarized here in order to present them in
the same form and using the same notation as in later chapters. In addition, this
chapter has emphasized the how-to aspects by providing a series of Matlab
exercises, which will be useful when creating simulations of the various parts of
a receiver.

For Further Reading

The intellectual background of the material presented here is often called signals
and systems. One of the most accessible books is

r J. H. McClellan, R. W. Schafer, and M. A. Yoder, Signal Processing First,

Pearson Prentice-Hall, 2003.
Other books provide greater depth and detail about the theory and uses of
Fourier transforms. We recommend these as both background and supplementary
reading:

r A. V. Oppenheim, A. S. Willsky, and S. H. Nawab, Signals and Systems, 2nd

edition, Prentice-Hall, 1997;

r S. Haykin and B. Van Veen, Signals and Systems, Wiley, 2002.

4

Assuming a six-ngered hand.

The Six Elements

57

There are also many wonderful books about digital signal processing, and these
provide both depth and detail about basic issues such as sampling and lter
design. Some of the best are the following:

r A. V. Oppenheim, R. W. Schafer, and J. R. Buck, Discrete-Time Signal Processing, Prentice-Hall, 1999;

r B. Porat, A Course in Digital Signal Processing, Wiley, 1997;

r S. Mitra, Digital Signal Processing: A Computer-Based Approach, McGrawHill, 2005.
Finally, since Matlab is fundamental to our presentation, it is worth mentioning
some books that describe the uses (and abuses) of the Matlab language.

r A. Gilat, MATLAB: An Introduction with Applications, Wiley, 2007;

r B. Littleeld and D. Hanselman, Mastering Matlab 7, Prentice-Hall, 2004.

Step 3: The Idealized System

The next step encompasses Chapters 4 through 9. This gives a closer look at
the idealized receiverhow things work when everything is just right: when the
timing is known, when the clocks run at exactly the right speed, when there are
no reections, diractions, or diusions of the electromagnetic waves. This step
also introduces a few Matlab tools that are needed to implement the digital
radio. The order in which topics are discussed is precisely the order in which
they appear in the receiver:
frequency
channel translation sampling
Chapter 4
Chapter 5
Chapter 6
receive
decision
equalization
decoding
ltering
device


 


Chapter 7
Chapter 8
Chapter 9 provides a complete (though idealized) software-dened digital radio
system.

Modeling Corruption

If every signal that went from here to there arrived at its intended receiver
unchanged, the life of a communications engineer would be easy. Unfortunately,
the path between here and there can be degraded in several ways, including
multipath interference, changing (fading) channel gains, interference from other
users, broadband noise, and narrowband interference.
This chapter begins by describing some of the funny things that can happen
to signals, some of which are diagrammed in Figure 4.1. More important than
locating the sources of the problems is xing them. The received signal can
be processed using linear lters to help reduce the interferences and to undo,
to some extent, the eects of the degradations. The central question is how to
specify lters that can successfully mitigate these problems, and answering this
requires a fairly detailed understanding of ltering. Thus, a discussion of linear
lters occupies the bulk of this chapter, which also provides a background for
other uses of lters throughout the receiver, such as the lowpass lters used in the
demodulators of Chapter 5, the pulse-shaping and matched lters of Chapter 11,
and the equalizing lters of Chapter 13.

4.1

When Bad Things Happen to Good Signals

The path from the transmitter to the receiver is not simple, as Figure 4.1 suggests. Before the signal reaches the receiver, it may be subject to a series of
strange events that can corrupt the signal and degrade the functioning of the
receiver. This section discusses ve kinds of corruption that are used throughout
the chapter to motivate and explain the various purposes that linear lters may
serve in the receiver.

60

Chapter 4. Modeling Corruption

Types of corruption

Transmitted
signal
Multipath

4.1.1

Changing
gain

Signals Broadband Narrowband

from
noise
noise
other
users
Received
signal
+
+
+

Figure 4.1 Sources of

corruption include
multipath interference,
changing channel gains,
interference from other
users, broadband noise,
and narrowband
interferences.

Other Users
Many dierent users must be able to broadcast at the same time. This requires
that there be a way for a receiver to separate the desired transmission from all
the others (for instance, to tune to a particular radio or TV station among a
large number that may be broadcasting simultaneously in the same geographical
region). One standard method is to allocate dierent frequency bands to each
user. This was called frequency-division multiplexing (FDM) in Chapter 2, and
was shown diagrammatically in Figure 2.3 on page 22. The signals from the
dierent users can be separated using a bandpass lter, as in Figure 2.4 on page
23. Of course, practical lters do not completely remove out-of-band signals, nor
do they pass in-band signals completely without distortions. Recall the three
lters in Figure 3.7 on page 48.

4.1.2

Broadband Noise
When the signal arrives at the receiver, it is small and must be amplied. While
it is possible to build high-gain ampliers, the noises and interferences will also
be amplied along with the signal. In particular, any noise in the amplier itself
will be increased. This is often called thermal noise and is usually modeled as
white (independent) broadband noise. Thermal noise is inherent in any electronic
component and is caused by small random motions of electrons, like the Brownian
motion of small particles suspended in water.
Such broadband noise is another reason why a bandpass lter is applied at
the front end of the receiver. By applying a suitable lter, the total power in
the noise (compared with the total power in the signal) can often be reduced.
Figure 4.2 shows the spectrum of the signal as a pair of triangles centered at the
carrier frequency fc with bandwidth 2B. The total power in the signal is the
area under the triangles. The spectrum of the noise is the at region, and its
power is the shaded area. After applying the bandpass lter, the power in the
signal remains (more or less) unchanged, while the power in the noise is greatly
reduced. Thus, the signal-to-noise ratio (SNR) improves.

Modeling Corruption

Power
in noise

fc B

fc + B

fc B

fc + B

(a)

(b)

fc B

4.1.3

fc + B

0
(c)

Power
in signal

fc B

fc + B

fc B

fc + B

Total
power in noise
is decreased
fc B

61

Power
in signal
is unchanged

Figure 4.2 The signal-to-noise

ratio is depicted graphically as

the ratio of the power of the
signal (the area under the
triangles) to the power in the
noise (the shaded area). After
the bandpass lter, the power
in the noise decreases, so the
SNR increases.

fc + B

Narrowband Noise
Noises are not always white; that is, the spectrum need not always be at. Stray
sine waves (and other signals with narrow spectra) may also impinge on the
receiver. These may be caused by errant transmitters that accidentally broadcast in the frequency range of the signal, or they may be harmonics of a lowerfrequency wave as it experiences nonlinear distortion. If these narrowband disturbances occur out of band, they will automatically be attenuated by the bandpass
lter just as if they were a component of the wideband noise. However, if they
occur in the frequency region of the signal, they decrease the SNR in proportion
to their power. Judicious use of a notch lter (one designed to remove just the
oending frequency) can be an eective tool.
Figure 4.3 shows the spectrum of the signal as the pair of triangles, along with
three narrowband interferers represented by the three pairs of spikes. After the
bandpass lter (BPF), the two pairs of out-of-band spikes are removed, but the
in-band pair remains. Applying a narrow notch lter tuned to the frequency
of the interferer allows its removal, although this cannot be done without also
aecting the signal somewhat.

4.1.4

Multipath Interference
In some situations, an electromagnetic wave can propagate directly from one
place to another. For instance, when a radio signal from a spacecraft is transmitted back to Earth, the vacuum of space guarantees that the wave will arrive
more or less intact (though greatly attenuated by distance). Often, however, the
wave reects, refracts, or diracts, and the signal arriving is quite dierent from
the one that was sent.

62

Chapter 4. Modeling Corruption

Signal plus three

narrowband
interferers

BPF

Removal of
out-of-band
interferers

Notch
filter

Figure 4.3 Three narrowband

interferers are shown in the top

gure (the three pairs of spikes).
The BPF cannot remove the
in-band interferer, though a
narrow notch lter can, at the
expense of changing the signal in
the region where the
narrowband noise occurred.

Removal of
in-band
interferers also
modifier signals

These distortions can be thought of as a combination of scaled and delayed

reections of the transmitted signal, which occur when there are dierent propagation paths from the transmitter to the receiver. Between two transmission
towers, for instance, the paths may include one along the line of sight, reections
from the atmosphere, reections from nearby hills, and bounces from a eld or
lake between the towers. For indoor digital TV reception, there are many (local)
time-varying reectors, including people in the receiving room, nearby vehicles,
and the buildings of an urban environment. Figure 4.4, for instance, shows multiple reections that arrive after bouncing o a cloud and o a mountain, and
others that are scattered by multiple bounces from nearby buildings.
The strength of the reections depends on the physical properties of the reecting surface, while the delay of the reections is primarily determined by the
length of the transmission path. Let s(t) be the transmitted signal. If N delays
are represented by 1 , 2 , . . . , N , and the strengths of the reections are
h1 , h2 , . . . , hN , then the received signal r(t) is
r(t) = h1 s(t 1 ) + h2 s(t 2 ) + + hN s(t N ).

(4.1)

As will become clear in Section 4.4, this model of the channel has the form of
a linear lter (since the expression on the right-hand side is a convolution of
the transmitted signal and the hi s). This is shown in part (a) of Figure 4.5.
Since this channel model is a linear lter, it can also be viewed in the frequency
domain, and part (b) shows its frequency response. When this is combined with
the BPF and the spectrum of the signal (shown in (c)), the result is the distorted
spectrum shown in (d).
What can be done?

Modeling Corruption

63

Cloud
Buildings

Figure 4.4 The received

Line of sigh

Receiver

signal may be a combination

of several copies of the
original transmitted signal,
each with a dierent
attenuation and delay.

Transmitter

Mountain

Transmitted
signal
(a)

Received
signal

h(t)
Fourier Transform

Figure 4.5 (a) The channel

model (4.1) as a lter. (b)

The frequency response of
the lter. (c) An idealized
BPF and the spectrum of the
signal. The product of (b)
and (c) gives (d), the
distorted spectrum at the
receiver.

H( f )

(b)

(c)

(d)

Frequency

If the kinds of distortions introduced by the channel are known (or can somehow be determined), then the bandpass lter at the receiver can be modied
in order to undo the eects of the channel. This can be seen most clearly in
the frequency domain, as in Figure 4.6. Observe that the BPF is shaped (part
(d)) to approximately invert the debilitating eects of the channel (part (a)) in
the frequency band of the signal and to remove all the out-of-band frequencies.
The resulting received signal spectrum (part (e)) is again a close copy of the
transmitted signal spectrum, in stark contrast to the received signal spectrum
in Figure 4.5 where no shaping was attempted.

64

Chapter 4. Modeling Corruption

Frequency
response of
channel
(a)
Spectrum
of
signal
(b)
Product of
(a) & (b)
(c)
BPF with
shaping
(d)
Product of
(c) & (d)
(e)
Figure 4.6 (a) The frequency response of the channel. (b) The spectrum of the signal.
(c) The product of (a) and (b), which is the spectrum of the received signal. (d) A
BPF lter that has been shaped to undo the eect of the channel. (e) The product of
(c) and (d), which combine to give a clean representation of the original spectrum of
the signal.

Thus, ltering in the receiver can be used to reshape the received signal within
the frequency band of the transmission as well as to remove unwanted out-ofband frequencies.

4.1.5

Fading
Another kind of corruption that a signal may encounter on its journey from the
transmitter to the receiver is called fading, where the frequency response of the
channel changes slowly over time. This may be caused because the transmission
path changes. For instance, a reection from a cloud might disappear when the
cloud dissipates, an additional reection might appear when a truck moves into a
narrow city street, or in a mobile device such as a cell phone the operator might
turn a corner and cause a large change in the local geometry of reections.
Fading may also occur when the transmitter and/or the receiver are moving.
The Doppler eect shifts the frequencies slightly, causing interferences that may
slowly change.
Such time-varying problems cannot be xed by a single xed lter; rather, the
lter must somehow compensate dierently at dierent times. This is an ideal
application for the adaptive elements of Section 3.7, though results from the
study of linear lters will be crucial in understanding how the time variations

Modeling Corruption

65

in the frequency response can be represented as time-varying coecients in the

lter that represents the channel.

4.2

Linear Systems: Linear Filters

Linear systems appear in many places in communication systems. The transmission channel is often modeled as a linear system as in (4.1). The bandpass
lters used in the front end to remove other users (and to remove noises) are
linear. Lowpass lters are crucial to the operation of the demodulators of Chapter 5. The equalizers of Chapter 13 are linear lters that are designed during the
operation of the receiver on the basis of certain characteristics of the received
signal.
Time-invariant linear systems can be described in any one of three equivalent
ways.

r The impulse response h(t) is a function of time that denes the output of a

linear system when the input is an impulse (or ) function. When the input
to the linear system is more complicated than a single impulse, the output
can be calculated from the impulse response via the convolution operator.
The frequency response H(f ) is a function of frequency that denes how the
spectrum of the input is changed into the spectrum of the output. The frequency response and the impulse response are intimately related: H(f ) is the
Fourier transform of h(t).
A linear dierence equation with constant coecients (such as (4.1)) shows
explicitly how the linear system can be implemented and can be useful in
assessing stability and performance.

This chapter describes the three representations of linear systems and shows
how they interrelate. The discussion begins by exploring the function, and then
showing how it is used to dene the impulse response. The convolution property
of the Fourier transform then shows that the transform of the impulse response
describes how the system behaves in terms of the input and output spectra, and
so it is called the frequency response. The nal step is to show how the action
of the linear system can be redescribed in the time domain as a dierence (or as
a dierential) equation. This is postponed to Chapter 7, and is also discussed in
some detail in Appendix F.

4.3

The Delta Function

One way to see how a system behaves is to kick it and see how it responds.
Some systems react sluggishly, barely moving away from their resting state,
while others respond quickly and vigorously. Dening exactly what is meant
mathematically by a kick is trickier than it seems because the kick must occur

66

Chapter 4. Modeling Corruption

over a very short amount of time, yet must be energetic in order to have any
eect. This section denes the impulse (or delta) function (t), which is a useful
kick for the study of linear systems.
The criterion that the impulse be energetic is translated to the mathematical
statement that its integral over all time must be nonzero, and it is typically
scaled to unity, that is,

(t)dt = 1.
(4.2)

The criterion that it occur over a very short time span is translated to the
statement that, for every positive ,

0 t < ,
(t) =
(4.3)
0 t > .
Thus, the impulse (t) is explicitly dened to be equal to zero for all t = 0. On
the other hand, (t) is implicitly dened when t = 0 by the requirement that its
integral be unity. Together, these guarantee that (t) is no ordinary function.1
The most important consequence of the denitions (4.2) and (4.3) is the sifting
property

w(t)(t t0 )dt = w(t)|t=t0 = w(t0 ),
(4.4)

which says that the delta function picks out the value of the function w(t) from
under the integral at exactly the time when the argument of the function is
zero, that is, when t = t0 . To see this, observe that (t t0 ) is zero whenever
t = t0 , and hence w(t)(t t0 ) is zero whenever t = t0 . Thus,


w(t)(t t0 )dt =
w(t0 )(t t0 )dt

(t t0 )dt = w(t0 ) 1 = w(t0 ).

= w(t0 )

Sometimes it is helpful to think of the impulse as a limit. For instance, dene

the rectangular pulse of width 1/n and height n by

0 t < 1/(2n),
n (t) = n 1/(2n) t 1/(2n),

0 t > 1/(2n).
Then (t) = limn n (t) fullls both criteria (4.2) and (4.3). Informally, it is not
unreasonable to think of (t) as being zero everywhere except at t = 0, where it is
innite. While it is not really possible to plot the delta function (t t0 ), it can
be represented in graphical form as zero everywhere except for an up-pointing
arrow at t0 . When the function is scaled by a constant, the value of the constant
is often placed in parentheses near the arrowhead. Sometimes, when the constant
1

The impulse is called a distribution and is the subject of considerable mathematical investigation.

Modeling Corruption

67

(3)
Figure 4.7 The function w(t) = (t + 10)

2(t + 1) + 3(t 5) consisting of three

weighted functions is represented
graphically as three weighted arrows at
t = 10, 1, 5, weighted by the
appropriate constants.

(1)
10

10

(2)

is negative, the arrow is drawn pointing down. For instance, Figure 4.7 shows a
graphical representation of the function w(t) = (t + 10) 2(t + 1) + 3(t 5).
What is the spectrum (Fourier transform) of (t)? This can be calculated
directly from the denition by replacing w(t) in (2.1) with (t):

(t)ej2f t dt.
(4.5)
F {(t)} =

Apply the sifting property (4.4) with w(t) = ej2f t and t0 = 0. Thus F {(t)} =
ej2f t |t=0 = 1.
Alternatively, suppose that is a function of frequency, that is, a spike at
zero frequency. The corresponding time-domain function can be calculated analogously using the denition of the inverse Fourier transform, that is, by substituting (f ) for W (f ) in (A.16) and integrating:

(f )ej2f t df = ej2f t |f =0 = 1.
F 1 {(f )} =

Thus a spike at frequency zero is a DC signal (a constant) in time.

The discrete time counterpart of (t) is the (discrete) delta function

1 k = 0,
[k] =
0 k = 0.
While there are a few subtle dierences between (t) and [k], for the most part
(t) and [k] act analogously. For example, the program specdelta.m calculates
the spectrum of the (discrete) delta function.
Listing 4.1. specdelta.m plots the spectrum of a delta function
time =2;
Ts =1/100;
t=Ts : Ts : time ;
x=zeros ( s i z e ( t ) ) ;
x(1)=1;
p l o t s p e c ( x , Ts )

%
%
%
%
%
%

l e n g t h o f time
time i n t e r v a l between s a m p l es
c r e a t e time v e c t o r
create si gnal of a l l zeros
delta function
draw waveform and spectrum

The output of specdelta.m is shown in Figure 4.8. As expected from (4.5),

the magnitude spectrum of the delta function is equal to 1 at all frequencies.
Exercise 4.1. Calculate the Fourier transform of (t t0 ) from the denition.
Now calculate it using the time-shift property (A.37). Are they the same? Hint:
they had better be.

Amplitude

Chapter 4. Modeling Corruption

Magnitude

68

0
0

0.2

0.4 0.6

0.8 1 1.2 1.4

Seconds

1.6 1.8

at time 0 has a magnitude spectrum

equal to 1 for all frequencies.

1
0

40 30 20 10 0 10
Frequency

20

Figure 4.8 A (discrete) delta function

30

40

Exercise 4.2. Use the denition of the IFT (A.16) to show that
(f f0 ) ej2f0 t .

Exercise 4.3. Mimic the code in specdelta.m to nd the magnitude spectrum

of the discrete delta function in the following cases.
a. The delta does not occur at the start of x. Try x[10]=1, x[100]=1, and
x[110]=1. How do the spectra dier? Can you use the time-shift property
(A.37) to explain what you see?
b. The delta changes magnitude x. Try x[1]=10, x[10]=3, and x[110]=0.1. How
do the spectra dier? Can you use the linearity property (A.31) to explain
what you see?
Exercise 4.4. Modify the code in specdelta.m to nd the phase spectrum of a
signal that consists of a delta function when the nonzero term is located at the
start (x(1)=1), in the middle (x(100)=1), and at the end (x(200)=1).
Exercise 4.5. Mimic the code in specdelta.m to nd the spectrum of a train of
equally spaced pulses. For instance, x(1:20:end)=1 spaces the pulses 20 samples
apart, and x(1:25:end)=1 places the pulses 25 samples apart.
a. Can you predict how far apart the resulting pulses in the spectrum will be?
b. Show that


1 
(t kTs )
(f nfs ),
(4.6)
Ts n=
k=

where fs = 1/Ts. Hint: let w(t) = 1 in (A.27) and (A.28).

c. Now can you predict how far apart the pulses in the spectrum are? Your
answer should be in terms of how far apart the pulses are in the time signal.
In Section 3.2, the spectrum of a sinusoid was shown to consist of two symmetric spikes in the frequency domain (recall Figure 3.5 on page 45). The next
example shows why this is true by explicitly taking the Fourier transform to nd
the spectrum of a sinusoid.

Modeling Corruption

69

|W( f )|
(A/2)

f0

Figure 4.9 The magnitude spectrum of a

(A/2)

f0

sinusoid with frequency f0 and amplitude A

contains two -function spikes, one at f = f0
and the other at f = f0 .

Example 4.1. Let w(t) = A sin(2f0 t), and use Eulers identity (A.3) to rewrite
w(t) as
w(t) =


A  j2f0 t
e
ej2f0 t .
2j

Applying the linearity property (A.31) and the result of Exercise 4.2 gives

A 
F {ej2f0 t } F{ej2f0 t }
2j
A
= j [(f f0 ) + (f + f0 )] .
2

F {w(t)} =

(4.7)

Thus, the spectrum of a sine wave is a pair of functions with opposite signs,
located symmetrically about zero frequency. The corresponding magnitude spectrum, shown in Figure 4.9, is at the heart of one important interpretation of the
Fourier transform: it shows the frequency content of any signal by displaying
which frequencies are present in (and absent from) the waveform. For example,
Figure 4.10(a) shows the magnitude spectrum W (f ) of a real-valued signal w(t).
This can be interpreted as saying that w(t) contains (or is made up of) all the
frequencies up to B Hz, and that it contains no sinusoids with higher frequency.
Similarly, the modulated signal s(t) in Figure 4.10(b) contains all positive frequencies between fc B and fc + B, and no others.
Note that the Fourier transform in (4.7) is purely imaginary, as it must be
because w(t) is odd (see (A.37)). The phase spectrum is a at line at 90
because of the factor j.
Exercise 4.6. What is the magnitude spectrum of sin(2f0 t + )? Hint: use the
frequency-shift property (A.34). Show that the spectrum of cos(2f0 t) is 12 ((f
f0 ) + (f + f0 )). Compare this analytical result with the numerical results from
Exercise 3.6.

Exercise 4.7. Let wi (t) = ai sin(2fi t) for i = 1, 2, 3. Without doing any calculations, write down the spectrum of v(t) = w1 (t) + w2 (t) + w3 (t). Hint: use
linearity. Graph the magnitude spectrum of v(t) in the same manner as in Figure 4.9. Verify your results with a simulation mimicking that in Exercise 3.7.

70

Chapter 4. Modeling Corruption

|W( f )|
1

|S( f )|
1/2

+B

(a)

fc B

fc

fc + B

fc B

fc

fc + B

(b)

Figure 4.10 The magnitude spectrum of a message signal w(t) is shown in (a). When

w(t) is modulated by a cosine at frequency fc , the spectrum of the resulting signal

s(t) = w(t) cos(2fc t + ) is shown in (b).

Exercise 4.8. Let W (f ) = sin(2f t0 ). What is the corresponding time function?

4.4

Convolution in Time: Its What Linear Systems Do

Suppose that a system has impulse response h(t), and that the input consists of
a sum of three impulses occurring at times t0 , t1 , and t2 , with amplitudes a0 ,
a1 , and a2 (for example, the signal w(t) of Figure 4.7). Because of the linearity
of the Fourier transform, property (A.31), the output is a superposition of the
outputs due to each of the input pulses. The output due to the rst impulse is
a0 h(t t0 ), which is the impulse response scaled by the size of the input and
shifted to begin when the rst input pulse arrives. Similarly, the outputs to the
second and third input impulses are a1 h(t t1 ) and a2 h(t t2 ), respectively,
and the complete output is the sum a0 h(t t0 ) + a1 h(t t1 ) + a2 h(t t2 ).
Now suppose that the input is a continuous function x(t). At any time instant
, the input can be thought of as consisting of an impulse scaled by the amplitude
x(), and the corresponding output will be x()h(t ), which is the impulse
response scaled by the size of the input and shifted to begin at time . The
complete output is then given by integrating over all :

y(t) =
x()h(t )d x(t) h(t).
(4.8)

This integral denes the convolution operator and provides a way of nding
the output y(t) of any linear system, given its impulse response h(t) and the
input x(t).
Matlab has several functions that simplify the numerical evaluation of convolutions. The most obvious of these is conv, which is used in convolex.m to
calculate the convolution of an input x (consisting of two delta functions at times
t = 1 and t = 3) and a system with impulse response h that is an exponential
pulse. The convolution gives the output of the system.
Listing 4.2. convolex.m example of numerical convolution
Ts =1/100; time =10;
t =0:Ts : time ;
h=exp( t ) ;
x=zeros ( s i z e ( t ) ) ;

%
%
%
%

s a m p l i n g i n t e r v a l and t o t a l time
c r e a t e time v e c t o r
def i ne impulse response
i n p u t = sum o f two d e l t a f u n c t i o n s

Modeling Corruption

71

Input

3
2
1

Impulse response

0
1

Figure 4.11 The convolution

of the input (the top plot)
with the impulse response of
the system (the middle plot)
gives the output in the
bottom plot.

0.5
0

Output

3
2
1
0
0

3
4
5
Time in seconds

x (1 / Ts )=3; x (3 / Ts )=2;
% a t t i m e s t=1 and t=3
y=conv ( h , x ) ;
% do c o n v o l u t i o n
subplot ( 3 , 1 , 1 ) , plot ( t , x )
% and p l o t
subplot ( 3 , 1 , 2 ) , plot ( t , h )
subplot ( 3 , 1 , 3 ) , plot ( t , y ( 1 : length ( t ) ) )

Figure 4.11 shows the input to the system in the top plot, the impulse response
in the middle plot, and the output of the system in the bottom plot. Nothing
happens before time t = 1, and the output is zero. When the rst spike occurs,
the system responds by jumping to 3 and then decaying slowly at a rate dictated
by the shape of h(t). The decay continues smoothly until time t = 3, when the
second spike enters. At this point, the output jumps up by 2, and is the sum of
the response to the second spike, plus the remainder of the response to the rst
spike. Since there are no more inputs, the output slowly dies away.
Exercise 4.9. Suppose that the impulse response h(t) of a linear system is the
exponential pulse
 t
t 0,
e
(4.9)
h(t) =
0
t < 0.
Suppose that the input to the system is 3(t 1) + 2(t 3). Use the denition
of convolution (4.8) to show that the output is 3h(t 1) + 2h(t 3), where
 t+t
0
e
t t0 ,
h(t t0 ) =
0
t < t0 .
How does your answer compare with Figure 4.11?

72

Chapter 4. Modeling Corruption

Exercise 4.10. Suppose that a system has an impulse response that is an exponential pulse. Mimic the code in convolex.m to nd its output when the input
is a white noise (recall specnoise.m on page 42).
Exercise 4.11. Mimic the code in convolex.m to nd the output of a system
when the input is an exponential pulse and the impulse response is a sum of two
delta functions at times t = 1 and t = 3.
The next two problems show that linear lters commute with dierentiation, and
with each other.
Exercise 4.12. Use the denition to show that convolution is commutative (i.e.,
that w1 (t) w2 (t) = w2 (t) w1 (t)). Hint: apply the change of variables = t
in (4.8).
Exercise 4.13. Suppose a lter has impulse response h(t). When the input is
x(t), the output is y(t). If the input is xd (t) = x(t)/t, the output is yd (t). Show
that yd (t) is the derivative of y(t). Hint: use (4.8) and the result of Exercise 4.12.
Exercise 4.14. Let w(t) = (t) be the rectangular pulse of (2.8). What is
w(t) w(t)? Hint: a pulse shaped like a triangle.
Exercise 4.15. Redo Exercise 4.14 numerically by suitably modifying
convolex.m. Let T = 1.5 s.
Exercise 4.16. Suppose that a system has an impulse response that is a sinc
function, as shown in Figure 2.11, and that the input to the system is a white
noise (as in specnoise.m on page 42).
a. Mimic convolex.m to numerically nd the output.
b. Plot the spectrum of the input and the spectrum of the output (using
plotspec.m). What kind of lter would you call this?

4.5

Convolution Multiplication
While the convolution operator (4.8) describes mathematically how a linear system acts on a given input, time-domain approaches are often not particularly
revealing about the general behavior of the system. Who would guess, for instance
in Exercises 4.11 and 4.16, that convolution with exponentials and sinc functions
would act like lowpass lters? By working in the frequency domain, however, the
convolution operator is transformed into a simpler point-by-point multiplication,
and the generic behavior of the system becomes clearer.
The rst step is to understand the relationship between convolution in time
and multiplication in frequency. Suppose that the two time signals w1 (t) and

Modeling Corruption

73

w2 (t) have Fourier transforms W1 (f ) and W2 (f ). Then,

F {w1 (t) w2 (t)} = W1 (f )W2 (f ).

(4.10)

To justify this property, begin with the denition of the Fourier transform (2.1)
and apply the denition of convolution (4.8) to obtain

F {w1 (t) w2 (t)} =
w1 (t) w2 (t)ej2f t dt
t=

 
=
w1 ()w2 (t )d ej2f t dt.
t=

Reversing the order of integration and using the time-shift property (A.37) produces



j2f t
w1 ()
w2 (t )e
dt d
F {w1 (t) w2 (t)} =
t=
=



=
w1 () W2 (f )ej2f d
=

= W2 (f )
w1 ()ej2f d = W1 (f )W2 (f ).
=

Thus, convolution in the time domain is the same as multiplication in the frequency domain. See (A.40).
The companion to the convolution property is the multiplication property,
which says that multiplication in the time domain is equivalent to convolution
in the frequency domain (see (A.41)); that is,

W1 ()W2 (f )d.
(4.11)
F {w1 (t)w2 (t)} = W1 (f ) W2 (f ) =

The usefulness of these convolution properties is apparent when applying them

to linear systems. Suppose that H(f ) is the Fourier transform of the impulse
response h(t). Suppose that X(f ) is the Fourier transform of the input x(t) that
is applied to the system. Then (4.8) and (4.10) show that the Fourier transform
of the output is exactly equal to the product of the transforms of the input and
the impulse response, that is,
Y (f ) = F {y(t)} = F {x(t) h(t)} = F {h(t)}F{x(t)} = H(f )X(f ).
This can be rearranged to solve for
H(f ) =

Y (f )
,
X(f )

(4.12)

which is called the frequency response of the system because it shows, for each
frequency f , how the system responds. For instance, suppose that H(f1 ) = 3 at
some frequency f1 . Then, whenever a sinusoid of frequency f1 is input into the
system, it will be amplied by a factor of 3. Alternatively, suppose that H(f2 ) = 0
at some frequency f2 . Then whenever a sinusoid of frequency f2 is input into

74

Chapter 4. Modeling Corruption

the system, it is removed from the output (because it has been multiplied by a
factor of 0).
The frequency response shows how the system treats inputs containing various
frequencies. In fact, this property was already used repeatedly in Chapter 1 when
drawing curves that describe the behavior of lowpass and bandpass lters. For
example, the lters of Figures 2.4, 2.5, and 2.6 are used to remove unwanted
frequencies from the communications system. In each of these cases, the plot of
the frequency response describes concretely and concisely how the system (or
lter) aects the input, and how the frequency content of the output relates to
that of the input. Sometimes, the frequency response H(f ) is called the transfer
function of the system, since it transfers the input x(t) (with transform X(f ))
into the output y(t) (with transform Y (f )).
Thus, the impulse response describes how a system behaves directly in time,
while the frequency response describes how it behaves in frequency. The two
descriptions are intimately related because the frequency response is the Fourier
transform of the impulse response. This will be used repeatedly in Section 7.2
to design lters for the manipulation (augmentation or removal) of specied
frequencies.

Example 4.2. In Exercise 4.16, a system was dened to have an impulse response
that is a sinc function. The Fourier transform of a sinc function in time is a rect
function in frequency (A.22). Hence, the frequency response of the system is a
rectangle that passes all frequencies below fc = 1/T and removes all frequencies
above (i.e., the system is a lowpass lter).
Matlab can help to visualize the relationship between the impulse response
and the frequency response. For instance, the system in convolex.m is dened
via its impulse response, which is a decaying exponential. Figure 4.11 shows
its output when the input is a simple sum of delta functions, and Exercise 4.10
explores the output when the input is a white noise. In freqresp.m, the behavior
of this system is explained by looking at its frequency response.
Listing 4.3. freqresp.m numerical example of impulse and frequency response
Ts =1/100; time =10;
t =0:Ts : time ;
h=exp( t ) ;
p l o t s p e c ( h , Ts )

%
%
%
%

s a m p l i n g i n t e r v a l and t o t a l time
c r e a t e time v e c t o r
def i ne impulse response
f i n d and p l o t f r e q u e n c y r e s p o n s e

The output of freqresp.m is shown in Figure 4.12. The frequency response

of the system (which is just the magnitude spectrum of the impulse response)
is found using plotspec.m. In this case, the frequency response amplies low
frequencies and attenuates other frequencies more as the frequency increases.
This explains, for instance, why the output of the convolution in Exercise 4.10
contained (primarily) lower frequencies, as evidenced by the slower undulations
in time.

Modeling Corruption

75

Amplitude

Figure 4.12 The action of a

6
Seconds

10

Magnitude

100

40 30

20

10
0
10
Frequency

20

30

system in time is dened by its

impulse response (in the top
plot). The action of the system
in frequency is dened by its
frequency response (in the
bottom plot), a kind of lowpass
lter.

40

Exercise 4.17. Suppose a system has an impulse response that is a sinc function. Using freqresp.m, nd the frequency response of the system. What kind
of lter does this represent? Hint: center the sinc in time; for instance, use
h=sinc(10*(t-time/2)).
Exercise 4.18. Suppose a system has an impulse response that is a sin function. Using freqresp.m, nd the frequency response of the system. What kind
of lter does this represent? Can you predict the relationship between the frequency of the sine wave and the location of the peaks in the spectrum? Hint: try
h=sin(25*t).
Exercise 4.19. Create a simulation (analogous to convolex.m) that inputs
white noise into a system with an impulse response that is a sinc function (as in
Exercise 4.17). Calculate the spectra of the input and output using plotspec.m.
Verify that the system behaves as suggested by the frequency response in Exercise 4.17.
Exercise 4.20. Create a simulation (analogous to convolex.m) that inputs
white noise into a system with an impulse response that is a sin function (as in
Exercise 4.18). Calculate the spectra of the input and output using plotspec.m.
Verify that the system behaves as suggested by the frequency response in Exercise 4.18.
So far, Section 4.5 has emphasized the idea of nding the frequency response of
a system as a way to understand its behavior. Reversing things suggests another
use. Suppose it was necessary to build a lter with some special characteristic
in the frequency domain (for instance, in order to accomplish one of the goals
of bandpass ltering in Section 4.1). It is easy to specify the lter in the frequency domain. Its impulse response can then be found by taking the inverse
Fourier transform, and the lter can be implemented using convolution. Thus,

76

Chapter 4. Modeling Corruption

the relationship between impulse response and frequency response can be used
both to study and to design systems.
In general, this method of designing lters is not optimal (in the sense that
other design methods can lead to more ecient designs), but it does show clearly
what the lter is doing, and why. Whatever the design procedure, the representation of the lter in the time domain and its representation in the frequency
domain are related by nothing more than a Fourier transform.

4.6

Improving SNR
Section 4.1 described several kinds of corruption that a signal may encounter
as it travels from the transmitter to the receiver. This section shows how linear
lters can help. Perhaps the simplest way a linear bandpass lter can be used
is to remove broadband noise from a signal. (Recall Section 4.1.2 and especially
Figure 4.2.)
A common way to quantify noise is the signal-to-noise ratio (SNR), which is
the ratio of the power of the signal to the power of the noise at a given point in
the system. If the SNR at one point is larger than the SNR at another point, the
performance is better at the rst point because there is more signal in comparison
with the amount of noise. For example, consider the SNR at the input and output
of a BPF as shown in Figure 4.13. The signal at the input (r(t) in part (a)) is
composed of the message signal x(t) and the noise signal n(t), and the SNR at
the input is therefore
SNRinput =

power in x(t)
.
power in n(t)

Similarly, the output y(t) is composed of a ltered version of the message (yx (t)
in part (b)) and a ltered version of the noise (yn (t) in part (b)). The SNR at
the output can therefore be calculated as
power in yx (t)
SNRoutput =
.
power in yn (t)
Observe that the SNR at the output cannot be calculated directly from y(t) (since
the two components are scrambled together). But, since the lter is linear,
y(t) = BPF{x(t) + n(t)} = BPF{x(t)} + BPF{n(t)} = yx + yn ,
which eectively shows the equivalence of parts (a) and (b) of Figure 4.13.
The Matlab program improvesnr.m explores this scenario concretely. The
signal x is a bandlimited signal, containing only frequencies between 3000 and
4000 Hz. This is corrupted by a broadband noise n (perhaps caused by an
internally generated thermal noise) to form the received signal. The SNR of this
input snrinp is calculated as the ratio of the power of the signal x to the power
of the noise n. The output of the BPF at the receiver is y, which is calculated

77

Modeling Corruption

n(t)

n(t)
x(t)

BPF
yn(t)
r(t)

y(t)

x(t)

yx(t)
BPF

BPF

y(t)
+

(b)

(a)

Figure 4.13 Two equivalent ways to draw the same system. In part (a) it is easy to
calculate the SNR at the input, while the alternative form (b) allows easy calculation
of the SNR at the output of the BPF.

as a BPF version of x+n. The BPF is created using the firpm command just
like the bandpass lter in filternoise.m on page 47. To calculate the SNR of
y, however, the code also implements the system in the alternative form of part
(b) of Figure 4.13. Thus, yx and yn represent the signal x ltered through the
BPF and the noise n passed through the same BPF. The SNR at the output is
then the ratio of the power in yx to the power in yn, which is calculated using
the function pow.m available on the website.
Listing 4.4. improvesnr.m using a linear lter to improve SNR
time =3; Ts =1/20000;
f r e q s =[0 0 . 2 9 0 . 3 0 . 4 0 . 4 1 1 ] ;
amps=[0 0 1 1 0 0 ] ;
b=f i r p m ( 1 0 0 , f r e q s , amps ) ;
n =0.25 * randn ( 1 , time /Ts ) ;
x=f i l t e r ( b , 1 , 2 * randn ( 1 , time /Ts ) ) ;
y=f i l t e r ( b , 1 , x+n ) ;
yx=f i l t e r ( b , 1 , x ) ;
yn=f i l t e r ( b , 1 , n ) ;
z=yx+yn ;
d i f f z y=max( abs ( zy ) )
s n r i n p=pow ( x )/ pow ( n )
s n r o u t=pow ( yx )/ pow ( yn )

%
%
%
%
%
%
%
%
%
%
%
%
%

time and s a m p l i n g i n t e r v a l
f i l t e r design , bandlimited
. . . between 3K and 4K
BP f i l t e r
g e n e r a t e w h i te n o i s e s i g n a l
do th e f i l t e r i n g
( a ) f i l t e r th e s i g n a l+n o i s e
or ( b) f i l t e r s i g n a l
. . . and n o i s e s e p a r a t e l y
add them
and make s u r e y = z
SNR a t i n p u t
SNR a t o u t p u t

Since the data generated in improvesnr.m are random, the numbers are
slightly dierent each time the program is run. Using the default values, the
SNR at the input is about 7.8, while the SNR at the output is about 61. This is
certainly a noticeable improvement. The variable diffzy shows the largest difference between the two ways of calculating the output (that is, between parts
(a) and (b) of Figure 4.13). This is on the order of 1015 , which is eectively
the numerical resolution of Matlab calculations, indicating that the two are
(eectively) the same.
Figure 4.14 plots the spectra of the input and the output of a typical run
of improvesnr.m. Observe the large noise oor in the left plot, and how this is
reduced by passage through the BPF. Observe also that the signal is still changed
by the noise in the passband between 3000 and 4000 Hz, since the BPF has no
eect there.

78

Chapter 4. Modeling Corruption

Magnitude spectrum of signal plus noise

Magnitude spectrum after filtering

Figure 4.14 The spectrum of the input to the BPF is shown in the left plot. The

spectrum of the output is shown on the right. The overall improvement in SNR is
clear.

The program improvesnr.m can be thought of as a simulation of the eect of

having a BPF at the receiver for the purposes of improving the SNR when the
signal is corrupted by broadband noise, as was described in Section 4.1.2. The
following problems ask you to mimic the code in improvesnr.m to simulate the
benet of applying lters to the other problems presented in Section 4.1.
Exercise 4.21. Suppose that the noise in improvesnr.m is replaced with narrowband noise (as discussed in Section 4.1.3). Investigate the improvements in
SNR
a. when the narrowband interference occurs outside the 3000 to 4000 Hz passband,
b. when the narrowband interference occurs inside the 3000 to 4000 Hz passband.
Exercise 4.22. Suppose that the noise in improvesnr.m is replaced with other
users who occupy dierent frequency bands (as discussed in Section 4.1.1). Are
there improvements in the SNR?
Exercise 4.23. Consider the interference between two users x1 and x2 occupying
the same frequency band as shown in Figure 4.15. The phases of the two mixers
at the transmitter are unequal with 5/6 > | | > /3. The lowpass lter
(LPF) has a cuto frequency of fc , a passband gain of 1.5, a stopband gain of
zero, and zero phase at zero frequency.
cos(2fct+)
receiver

x1(t)

y(t)
transmitter
x2(t)

LPF

cos(2fct+)
cos(2fct+)

Figure 4.15 The

multiuser transmission
system of Exercise
4.23.

Modeling Corruption

79

a. For this system Y (f ) = c1 X1 (f ) + c2 X2 (f ). Determine c1 and c2 as functions

of fc , , , and .
b. With set to maximize (c1 /c2 )2 , nd c1 and c2 as functions of fc , , and .
c. With set to minimize (c1 /c2 )2 , nd c1 and c2 as functions of fc , , and .
The other two problems posed in Section 4.1 were multipath interference and
fading. These require more sophisticated processing because the design of the
lters depends on the operating circumstances of the system. These situations
will be discussed in detail in Chapters 6 and 13.

For Further Reading

An early description of the linearity of communication channels can be found in

r P. A. Bello, Characterization of Randomly Time-Variant Linear Channels,

IEEE Transactions on Communication Systems, vol. 11, no. 4, pp 360393,
Dec. 1963.

Analog (De)modulation

Several parts of a communication system modulate the signal and change the
underlying frequency band in which the signal lies. These frequency changes
must be reversible; after processing, the receiver must be able to reconstruct (a
close approximation to) the transmitted signal.
The input message w(kT ) in Figure 5.1 is a discrete-time sequence drawn from
a nite alphabet. The ultimate output m(kT ) produced by the decision device
(or quantizer) is also discrete-time and is drawn from the same alphabet. If all
goes well and the message is transmitted, received, and decoded successfully, then
the output should be the same as the input, although there may be some delay
between the time of transmission and the time when the output is available.
Though the system is digital in terms of the message communicated and the
performance assessment, the middle of the system is inherently analog from the
(pulse-shaping) lter of the transmitter to the sampler at the receiver.
At the transmitter in Figure 5.1, the digital message has already been turned
into an analog signal by the pulse shaping (which was discussed briey in Section
2.10 and is considered in detail in Chapter 11). For ecient transmission, the
analog version of the message must be shifted in frequency, and this process
of changing frequencies is called modulation or upconversion. At the receiver,
the frequency must be shifted back down, and this is called demodulation or
downconversion. Sometimes the demodulation is done in one step (all analog) and
sometimes the demodulation proceeds in two steps, an analog downconversion to
the intermediate frequency and then a digital downconversion to the baseband.
This two-step procedure is shown in Figure 5.1.
There are many ways in which signals can be modulated. Perhaps the simplest is amplitude modulation (AM), which is discussed in two forms (large and
small carrier) in the next two sections. This is generalized to the simultaneous transmission of two signals using quadrature modulation in Section 5.3, and
it is shown that quadrature modulation uses bandwidth more eciently than
amplitude modulation. This gain in eciency can also be obtained using singlesideband and vestigial-sideband methods, which are discussed in the document

Analog (De)modulation

81

Interferers Noise
Analog
Pulse
upconversion
shaping
Message
w(kT) {3, 1, 1, 3}

Analog
conversion
to IF

Ts

Digital downconversion
to baseband
f

Decision

Reconstructed
message
m(kT ) {3, 1, 1, 3}

Carrier synchronization
Figure 5.1 A complete digital communication system has many parts. This chapter

focuses on the upconversion and the downconversion, which can done in many ways,
including large-carrier AM as in Section 5.1, suppressed-carrier AM as in Section 5.2,
and quadrature modulation as in Section 5.3.

titled Other Modulations available on the website. Demodulation can also be

accomplished using sampling, as discussed in Section 6.2, and amplitude modulation can also be accomplished with a simple squaring and ltering operation
as in Exercise 5.9.
Throughout, the chapter contains a series of exercises that prepare readers to
create their own modulation and demodulation routines in Matlab. These lie
at the heart of the software receiver that will be assembled in Chapters 9 and
15.

5.1

Amplitude Modulation with Large Carrier

Perhaps the simplest form of (analog) transmission system modulates the message signal by a high-frequency carrier in a two-step procedure: multiply the
message by the carrier, then add the product to the carrier. At the receiver, the
message can be demodulated by extracting the envelope of the received signal.
Consider the transmitted/modulated signal
v(t) = Ac w(t) cos(2fc t) + Ac cos(2fc t) = Ac (w(t) + 1) cos(2fc t)
diagrammed in Figure 5.2. The process of multiplying the signal in time by a
(co)sinusoid is called mixing. This can be rewritten in the frequency domain by
mimicking the development from (2.3) to (2.5) on page 20. Using the convolution
property of Fourier transforms (4.10), the transform of v(t) is
V (f ) = F {Ac (w(t) + 1) cos(2fc t)} = Ac F {(w(t) + 1)} F{cos(2fc t)}. (5.1)
The spectra of F {w(t) + 1} and |V (f )| are sketched in Figure 5.3 (a) and (b). The
vertical arrows in (b) represent the transform of the cosine carrier at frequency
fc (i.e., a pair of delta functions at fc ) and the scaling by Ac /2 is indicated
next to the arrowheads.
If w(t) 1, the envelope of v(t) is the same as w(t) and an envelope detector
can be used as a demodulator (envelopes are discussed in detail in Appendix C).
One way to nd the envelope of a signal is to lowpass lter the absolute value.

82

Chapter 5. Analog (De)modulation

w(t)

v(t)

v(t)

m(t)

|.|

LPF

envelope detector

Ac cos(2fct)

(b) Receiver

(a) Transmitter

Figure 5.2 A communications system using amplitude modulation with a large carrier.
In the transmitter (a), the message signal w(t) is modulated by a carrier wave at
frequency fc and then added to the carrier to give the transmitted signal v(t). In (b),
the received signal is passed through an envelope detector consisting of an
absolute-value nonlinearity followed by a lowpass lter. When all goes well, the
output m(t) of the receiver is approximately equal to the original message.

(a) |F{w(t) + 1}|

(Ac)

B +B

(b) |V( f )|
(Ac/2)

(Ac/2)

Figure 5.3 Spectra of the signals

fc

fc

(c) |F{|cos(2fct)|}|

4fc

2fc

2fc

4fc

2fc

4fc

in the large-carrier AM system

of Figure 5.2. Lowpass ltering
(d) gives a scaled version of (a).

(d) |F{|v(t)|}|

4fc

2fc

To see this analytically, observe that

F {|v(t)|} = F {|Ac (w(t) + 1) cos(2fc t)|}
= |Ac |F{|w(t) + 1|| cos(2fc t)|} = |Ac |F {w(t) + 1} F{| cos(2fc t)|},
where the absolute value can be removed from w(t) + 1 because w(t) + 1 > 0 (by
assumption). The spectrum of F {|cos(2fc t)|}, shown in Figure 5.3(c), may be
familiar from Exercise 3.23. Accordingly, F {|v(t)|} is the convolution shown in
Figure 5.3(d). Lowpass ltering this returns w(t) + 1, which is the envelope of
v(t) oset by the constant unity.

Analog (De)modulation

83

An example is given in the following Matlab program. The message signal

is a sinusoid with a drift in the DC oset, and the carrier wave is at a much
higher frequency.
Listing 5.1. AMlarge.m large-carrier AM demodulated with envelope
time = 0 . 3 3 ; Ts =1/10000;
t =0: Ts : time ; l e n t=length ( t ) ;
fm=20; f c =1000; c=cos (2 * pi * f c * t ) ;
w=10/ l e n t * [ 1 : l e n t ]+ cos (2 * pi * fm * t ) ;
v=c . * w+c ;
f b e =[0 0 . 0 5 0 . 1 1 ] ;
damps=[1 1 0 0 ] ; f l =100;
b=f i r p m ( f l , f b e , damps ) ;
envv =(pi / 2 ) * f i l t e r ( b , 1 , abs ( v ) ) ;

%
%
%
%
%
%

s a m p l i n g i n t e r v a l & time
d e f i n e a time v e c t o r
d e f i n e c a r r i e r at f r e q f c
c r e a t e message > 1
modulate w/ l a r g e c a r r i e r
LPF d e s i g n

% i m p u l s e r e s p o n s e o f LPF
% f i nd envelope

The output of this program is shown in Figure 5.4. The slowly increasing
sinusoidal message w(t) is modulated by the carrier c(t) at fc = 1000 Hz.
The heart of the modulation is the point-by-point multiplication of the message
and the carrier in the fth line. This product v(t) is shown in Figure 5.4(c). The
enveloping operation is accomplished by applying a lowpass lter to the real part
of 2v(t)ej2fc t (as discussed in Appendix C). This recovers the original message
signal, though it is oset by 1 and delayed by the linear lter.
Exercise 5.1. Using AMlarge.m, plot the spectrum of the message w(t), the
spectrum of the carrier c(t), and the spectrum of the received signal v(t). What
is the spectrum of the envelope? How close are your results to the theoretical
predictions in (5.1)?
Exercise 5.2. One of the advantages of transmissions using AM with a large
carrier is that there is no need to know the (exact) phase or frequency of the
transmitted signal. Verify this using AMlarge.m.
a. Change the phase of the transmitted signal; for instance, let c=cos(2*pi*
fc*t+phase) with phase=0.1, 0.5, pi/3, pi/2, pi, and verify that the
recovered envelope remains unchanged.
b. Change the frequency of the transmitted signal; for instance, let c=cos(2*
pi*(fc+g)*t) with g=10, -10, 100, -100, and verify that the recovered
envelope remains unchanged. Can g be too large?

Exercise 5.3. Create your own message signal w(t), and rerun AMlarge.m.
Repeat Exercise 5.1 with this new message. What dierences do you see?
Exercise 5.4. In AMlarge.m, verify that the original message w and the recovered
envelope envv are oset by 1, except at the end points where the lter does not
have enough data. Hint: the delay induced by the linear lter is approximately
fl/2.

Chapter 5. Analog (De)modulation

Amplitude

84

4
2
0

Amplitude

(a) message signal

1
0
1

Figure 5.4 An undulating

message (top) is modulated by

a carrier (b). The composite
signal is shown in (c), and the
output of an envelope detector
is shown in (d).

(b) carrier

Amplitude

Amplitude

4
2
0
2
(c) modulated signal

4
2
0
0

0.01

0.02

0.03

0.04
0.05 0.06
Seconds
(d) output of envelope detector

0.07

The principal advantage of transmission systems that use AM with a large

carrier is that exact synchronization is not needed; the phase and frequency of
the transmitter need not be known at the receiver, as was demonstrated in Exercise 5.2. This means that the receiver can be simpler than when synchronization
circuitry is required. The main disadvantage is that adding the carrier into the
signal increases the power needed for transmission but does not increase the
amount of useful information transmitted. Here is a clear engineering trade-o;
the value of the wasted signal strength must be balanced against the cost of the
receiver.

5.2

Amplitude Modulation with Suppressed Carrier

It is also possible to use AM without adding the carrier. Consider the transmitted/modulated signal
v(t) = Ac w(t)cos(2fc t)
diagrammed in Figure 5.5(a), in which the message w(t) is mixed with the cosine
carrier. Direct application of the frequency-shift property of Fourier transforms
(A.33) shows that the spectrum of the received signal is
V (f ) =

1
1
Ac W (f + fc ) + Ac W (f fc ).
2
2

As with AM with a large carrier, the upconverted signal v(t) for AM with a
suppressed carrier has twice the bandwidth of the original message signal. If the

Analog (De)modulation

w(t)

v(t)

v(t)

x(t)

85

m(t)
LPF

Ac cos(2fct)
(a) Transmitter

cos(2(fc + )t + )
(b) Receiver

Figure 5.5 A communications system using amplitude modulation with a suppressed

carrier. In the transmitter (a), the message signal w(t) is modulated by a carrier wave
at frequency fc to give the transmitted signal v(t). In (b), the received signal is
demodulated by a wave with frequency fc + and phase , and then lowpass ltered.
When all goes well, the output of the receiver m(t) is approximately equal to the
original message.

original message occupies the frequencies between B Hz, then the modulated
message has support between fc B and fc + B, a bandwidth of 2B. See Figure
4.10 on page 70.
As illustrated in (2.6) on page 25, the received signal can be demodulated by
mixing with a cosine that has the same frequency and phase as the modulating
cosine, and the original message can then be recovered by lowpass ltering. But,
as a practical matter, the frequency and phase of the modulating cosine (located
at the transmitter) can never be known exactly at the receiver.
Suppose that the frequency of the modulator is fc but that the frequency at
the receiver is fc + , for some small . Similarly, suppose that the phase of the
modulator is 0 but that the phase at the receiver is . Figure 5.5(b) shows this
downconverter, which can be described by
x(t) = v(t) cos(2(fc + )t + )

(5.2)

and
m(t) = LPF{x(t)},
where LPF represents a lowpass ltering of the demodulated signal x(t) in an
attempt to recover the message. Thus, the downconversion described in (5.2)
acknowledges that the receivers local oscillator might not have the same frequency or phase as the transmitters local oscillator. In practice, accurate a
priori information is available for the carrier frequency, but the (relative) phase
could be anything, since it depends on the distance between the transmitter and
the receiver as well as when the transmission begins. Because the frequencies are
high, the wavelengths are small and even small motions can change the phase
signicantly.
The remainder of this section investigates what happens when the frequency
and phase are not known exactly, that is, when or (or both and ) are
nonzero. Using the frequency-shift property of Fourier transforms (5.3) on x(t)

86

Chapter 5. Analog (De)modulation

in (5.2) produces the Fourier transform X(f )

Ac j
[e {W (f + fc (fc + )) + W (f fc (fc + ))}
4
+ ej {W (f + fc + (fc + )) + W (f fc + (fc + ))}]
Ac j
[e W (f ) + ej W (f 2fc )
=
4
+ ej W (f + 2fc + ) + ej W (f + )].

(5.3)

If there is no frequency oset (i.e., if = 0), then

X(f ) =

Ac j
[(e + ej )W (f ) + ej W (f 2fc ) + ej W (f + 2fc )].
4

Because cos(x) = (1/2)(ejx + ejx ) from (A.2), this can be rewritten


Ac
Ac  j
W (f )cos() +
e W (f 2fc ) + ej W (f + 2fc ) .
2
4
Thus, a perfect lowpass ltering of x(t) with cuto below 2fc removes the highfrequency portions of the signal near 2fc to produce
X(f ) =

Ac
w(t) cos().
(5.4)
2
The factor cos() attenuates the received signal (except for the special case when
= 0 2k for integers k). If were suciently close to 0 2k for some integer
k, then this would be tolerable. But there is no way to know the relative phase,
and hence cos() can assume any possible value within [1, 1]. The worst case
occurs as approaches /2, when the message is attenuated to zero! A scheme
for carrier-phase synchronization, which automatically tries to align the phase
of the cosine at the receiver with the phase at the transmitter, is vital. This is
discussed in detail in Chapter 10.
To continue the investigation, suppose that the carrier-phase oset is zero,
(i.e., = 0), but that the frequency oset is not. Then the spectrum of x(t)
from (5.3) is
m(t) =

X(f ) =

Ac
[W (f ) + W (f 2fc ) + W (f + 2fc + ) + W (f + )],
4

and the lowpass ltering of x(t) produces

Ac
[W (f ) + W (f + )] .
4
This is shown in Figure 5.6. Recognizing this spectrum as a frequency-shifted
version of w(t), it can be translated back into the time domain using (A.33) to
give
M (f ) =

Ac
w(t) cos(2t).
(5.5)
2
Instead of recovering the message w(t), the frequency oset causes the receiver
to recover a low-frequency amplitude-modulated version of it. This is bad with
m(t) =

Analog (De)modulation

|M( f )|

87

|W( f + )|
Figure 5.6 When there is a

Ac
4

|W( f )|

carrier-frequency oset in the receiver

oscillator, the two images of W () do
not align properly. Their sum is not
equal to (Ac /2)W (f ).

even a small carrier-frequency oset. While cos() in (5.4) is a xed scaling, cos(2t) in (5.5) is a time-varying scaling that will alternately recover
m(t) (when cos(2t) 1) and make recovery impossible (when cos(2t) 0).
Transmitters are typically expected to maintain suitable accuracy to a nominal
carrier-frequency setting known to the receiver. Ways of automatically tracking
(inevitable) small frequency deviations are discussed at length in Chapter 10.
The following code AM.m generates a message w(t) and modulates it with a
carrier at frequency fc . The demodulation is done with a cosine of frequency fc +
and a phase oset of . When = 0 and = 0, the output (a lowpass version
of the demodulated signal) is nearly identical to the original message, except for
the inevitable delay caused by the linear lter. Figure 5.7 shows four plots: the
message w(t) on top, followed by the upconverted signal v(t) = w(t)cos(2fc t),
followed in turn by the downconverted signal x(t). The lowpass-ltered version
is shown in the bottom plot; observe that it is nearly identical to the original
message, albeit with a slight delay.
Listing 5.2. AM.m suppressed carrier with (possible) freq and phase oset
time = 0 . 3 ; Ts =1/10000;
t=Ts : Ts : time ; l e n t=length ( t ) ;
fm=20; f c =1000; c=cos (2 * pi * f c * t ) ;
w=5/ l e n t * ( 1 : l e n t )+cos (2 * pi * fm * t ) ;
v=c . * w ;
gam=0; p h i =0;
c2=cos (2 * pi * ( f c+gam ) * t+p h i ) ;
x=v . * c2 ;
f b e =[0 0 . 1 0 . 2 1 ] ; damps=[1 1 0 0 ] ;
f l =100; b=f i r p m ( f l , f b e , damps ) ;
m=2* f i l t e r ( b , 1 , x ) ;

%
%
%
%
%
%
%
%
%
%
%

s a m p l i n g i n t e r v a l & time
d e f i n e a time v e c t o r
c a r r i e r at f r e q f c
c r e a t e message
modulate with c a r r i e r
f r e q & phase o f f s e t
c r e a t e c o s i n e f o r demod
demod r e c e i v e d s i g n a l
LPF d e s i g n
i m p u l s e r e s p o n s e o f LPF
LPF th e demodulated s i g n a l

Exercise 5.5. Using AM.m as a starting point, plot the spectra of w(t), v(t), x(t),
and m(t).
Exercise 5.6. Try dierent phase osets = [, /2, /3, /6, 0, /6,
/3, /2, ]. How well does the recovered message m(t) match the actual message
w(t)? For each case, what is the spectrum of m(t)?
Exercise 5.7. TRUE or FALSE: A small, xed phase oset in the receiver
demodulating AM with suppressed carrier produces an undesirable low-frequency
modulated version of the analog message.

Chapter 5. Analog (De)modulation

3
1
1
(a) message signal

Figure 5.7 The message signal

0
Amplitude

88

in the top frame is modulated

to produce the signal in the
second plot. Demodulation
gives the signal in the third
plot, and the LPF recovers
the original message (with
delay) in the bottom plot.

2
(b) message after modulation

3
1
1

(c) demodulated signal

3
1
1

0.02
0.04
0.06
0.08
(d) recovered message is a LPF applied to (c)

Addition
w(t)

+
A0 cos(2fct)

Square-law
device

()

Bandpass
filter
Centered
at fc

2A0w(t)cos(2fct)

Figure 5.8 The square-law

mixing transmitter of
Exercises 5.9 through
5.11.

Exercise 5.8. Try dierent frequency osets gam = [0.014, 0.1, 1.0, 10]. How
well does the recovered message m(t) match the actual message w(t)? For each
case, what is the spectrum of m(t)? Hint: look over more than just the rst 0.1 s
to see the eect.
Exercise 5.9. Consider the system shown in Figure 5.8. Show that the output
of the system is 2A0 w(t) cos(2fc t), as indicated.
Exercise 5.10. Create a Matlab routine to implement the square-law mixing
modulator of Figure 5.8.
a. Create a signal w(t) that has bandwidth 100 Hz.
b. Modulate the signal to 1000 Hz.
c. Demodulate using the AM demodulator from AM.m (to recover the original
w(t)).

89

Analog (De)modulation

|W( f )|
2

sin(2f4t) + cos(2f5t)

(a)
r(t)

x(t)
(c)
2000

(b)

300

300

w(t)

2000
x(t)

BPF
lower cutoff f2
cos(2f1t + 1) upper cutoff f3

BPF

(d)

r(t)
BPF

y(t)
LPF

lower cutoff f6
cutoff f9
upper cutoff f7 cos(2f8t + 2)

Figure 5.9 The transmission system for Exercise 5.12: (a) the magnitude spectrum of

the message, (b) the transmitter, (c) the channel, and (d) the receiver.

Exercise 5.11. Exercise 5.10 essentially creates a transmitter and receiver based
on the square-law modulator (rather than the more standard mixing modulator).
Using this system, do the following.
a. Show how the received signal degrades if the phase of the cosine wave is not
known exactly.
b. Show how the received signal degrades if the frequency of the cosine wave is
not exact.
c. Show how the received signal degrades if the bandpass lter is not centered
at the specied frequency.
Exercise 5.12. Consider the transmission system of Figure 5.9. The message
signal w(t) has the magnitude spectrum shown in part (a). The transmitter in
part (b) produces the transmitted signal x(t), which passes through the channel in part (c). The channel scales the signal and adds narrowband interferers
to create the received signal r(t). The transmitter and channel parameters are
1 = 0.3 radians, f1 = 24.1 kHz, f2 = 23.9 kHz, f3 = 27.5 kHz, f4 = 29.3 kHz,
and f5 = 22.6 kHz. The receiver processing r(t) is shown in Figure 5.9(d). All
bandpass and lowpass lters are considered ideal, with a gain of unity in the
passband and zero in the stopband.
a. Sketch |R(f )| for 30 kHz f 30 kHz. Clearly indicate the amplitudes and
frequencies of key points in the sketch.
b. Assume that 2 is chosen to maximize the magnitude of y(t) and reects the
value of 1 and the delays imposed by the two ideal bandpass lters that form
the received signal r(t). Select the receiver parameters f6 , f7 , f8 , and f9 , so
the receiver output y(t) is a scaled version of w(t).
Exercise 5.13. An analog baseband message signal w(t) has all energy between
B and B Hz. It is upconverted to the transmitted passband signal x(t) via AM

90

Chapter 5. Analog (De)modulation

with suppressed carrier

x(t) = w(t) cos(2fc t + c ),
where the carrier frequency fc > 10B. The channel is a pure delay and the
received signal r is r(t) = x(t d), where the delay d = nTc + Tc / is an integer
multiple n 0 of the carrier period Tc (= 1/fc) plus a fraction of Tc given by
> 1. The mixer at the receiver is perfectly synchronized to the transmitter so
that the mixer output y(t) is
y(t) = r(t) cos(2fc t + r ).
The receiver mixer phase need not match the transmitter mixer phase c . The
receiver lowpass lters y to produce
v(t) = LPF{y(t)},
where the lowpass lter is ideal with unity passband gain, linear passband phase
with zero phase at zero frequency, and cuto frequency 1.2B.
a. Write a formula for the receiver mixer output y(t) as a function of fc , c , d,
, r , and w(t) (without use of x, r, n, or Tc ).
b. Determine the amplitude of the minimum and maximum values of y(t) for
= 4.
c. For = 6, n = 42, c = 0.2 radians, and Tc = 20 s, determine the value of
r that maximizes the magnitude of the maximum and minimum values of
v(t).

5.3

Quadrature Modulation
In AM transmission, where the baseband signal and its modulated passband version are real-valued, the spectrum of the modulated signal has twice the bandwidth of the baseband signal. As pictured in Figure 4.10 on page 70, the spectrum
of the baseband signal is nonzero only for frequencies between B and B. After
modulation, the spectrum is nonzero in the interval [fc B, fc + B] and in
the interval [fc B, fc + B]. Thus the total width of frequencies occupied by the
modulated signal is twice that occupied by the baseband signal. This represents
a kind of ineciency or redundancy in the transmission. Quadrature modulation
provides one way of removing this redundancy by sending two messages in the
frequency ranges between [fc B, fc + B] and [fc B, fc + B], thus utilizing
the spectrum more eciently.
To see how this can work, suppose that there are two message streams m1 (t)
and m2 (t). Modulate one message with a cosine to create the in-phase signal,
and the other with (the negative of) a sine to form the quadrature signal. These

91

Analog (De)modulation

cos(2fct)

cos(2fct)

m1(t)

x1(t)

LPF

s1(t)

+
Transmitter
m2(t)

Receiver

sin(2fct)

x2(t)

LPF

s2(t)

Figure 5.10 In a quadrature modulation

system, two messages m1 (t) and m2 (t)
are modulated by two sinusoids of the
same frequency, sin(2fc t) and
cos(2fc t). The receiver then
demodulates twice and recovers the
original messages after lowpass ltering.

sin(2fct)

are summed1 to form

v(t) = m1 (t) cos(2fc t) m2 (t) sin(2fc t),

(5.6)

which is then transmitted. A receiver structure that can recover the two messages
is shown in Figure 5.10. The signal s1 (t) at the output of the receiver is intended
to recover the rst message m1 (t). Similarly, the signal s2 (t) at the output of the
receiver is intended to recover the (negative of the) second message m2 (t).
To examine the recovered signals s1 (t) and s2 (t) in Figure 5.10, rst evaluate
the signals before the lowpass ltering. Using the trigonometric identities (A.4)
and (A.8), x1 (t) becomes
x1 (t) = v(t) cos(2fc t) = m1 (t) cos2 (2fc t) m2 (t) sin(2fc t) cos(2fc t)
m1 (t)
m2 (t)
(1 + cos(4fc t))
sin(4fc t).
=
2
2
Lowpass ltering x1 (t) produces
m1 (t)
.
2
Similarly, x2 (t) can be rewritten using (A.5) and (A.8) as
s1 (t) =

x2 (t) = v(t) sin(2fc t) = m1 (t) cos(2fc t) sin(2fc t) m2 (t) sin2 (2fc t)

m1 (t)
m2 (t)
=
sin(4fc t)
(1 cos(4fc t)),
2
2
and lowpass ltering x2 (t) produces
m2 (t)
.
2
Thus, in the ideal situation in which the phases and frequencies of the modulation and the demodulation are identical, both messages can be recovered. But
if the frequencies and/or phases are not exact, then problems analogous to those
s2 (t) =

These are also sometimes modeled as the real and the imaginary parts of a single
complex-valued signal. This complex representation is explored more fully in Appendix
C.

92

Chapter 5. Analog (De)modulation

cos(2f1t)
x1(t)
x2(t)

x3(t)

x4(t)
BPF

x5(t)

x6(t)
LPF

Figure 5.11 The transmission

system of Exercise 5.16.

lower cutoff f3
cutoff f6
upper cutoff f4 cos(2f5t + )

cos(2f2t)

encountered with AM will occur in the quadrature modulation. For instance,

if the phase of (say) the demodulator x1 (t) is not correct, then there will be
some distortion or attenuation in s1 (t). However, problems in the demodulation
of s1 (t) may also cause problems in the demodulation of s2 (t). This is called
cross-interference between the two messages.
Exercise 5.14. Use AM.m as a starting point to create a quadrature modulation
system that implements the block diagram of Figure 5.10.
a. Examine the eect of a phase oset in the demodulating sinusoids of the
receiver, so that x1 (t) = v(t) cos(2fc t + ) and x2 (t) = v(t) sin(2fc t + )
for a variety of . Refer to Exercise 5.6.
b. Examine the eect of a frequency oset in the demodulating sinusoids of
the receiver, so that x1 (t) = v(t) cos(2(fc + )t) and x2 (t) = v(t) sin(2(fc +
)t) for a variety of . Refer to Exercise 5.8.
c. Conrm that a 1 phase error in the receiver oscillator corresponds to more
than 1% cross-interference.
Exercise 5.15. TRUE or FALSE: Quadrature amplitude modulation can combine two real, baseband messages of absolute bandwidth B in a radio-frequency
signal of absolute bandwidth B.
Exercise 5.16. Consider the scheme shown in Figure 5.11. The absolute bandwidth of the baseband signal x1 is 6 MHz and that of the baseband signal x2 (t)
is 4 MHz, f1 = 164 MHz, f2 = 154 MHz, f3 = 148 MHz, f4 = 160 MHz, f5 = 80
MHz, = /2, and f6 = 82 MHz.
a.
b.
c.
d.
e.

What
What
What
What
What

is
is
is
is
is

the
the
the
the
the

absolute bandwidth of x3 (t)?

absolute bandwidth of x5 (t)?
absolute bandwidth of x6 (t)?
maximum frequency in x3 (t)?
maximum frequency in x5 (t)?

Thus the ineciency of real-valued double-sided AM transmission can be

reduced using complex-valued quadrature modulation, which recaptures the lost
bandwidth by sending two messages simultaneously. For simplicity and clarity,

93

Analog (De)modulation

the bulk of Software Receiver Design focuses on the real PAM case and the
complex-valued quadrature case is postponed until Chapter 16. There are also
other ways of recapturing the lost bandwidth: both single-sideband and vestigialsideband (discussed in the document Other Modulations on the website) send a
single message, but use only half the bandwidth.

5.4

Injection to Intermediate Frequency

All the modulators and demodulators discussed in the previous sections downconvert to baseband in a single step, that is, the spectrum of the received signal
is shifted by mixing with a cosine of frequency fc that matches the transmission
frequency fc . As suggested in Section 2.8, it is also possible to downconvert to
some desired intermediate frequency (IF) fI (as depicted in Figure 2.9), and to
then later downconvert to baseband by mixing with a cosine of the intermediate
frequency fI . There are several advantages to such a two-step procedure:

r all frequency bands can be downconverted to the same IF, which allows use
of standardized ampliers, modulators, and lters on the IF signals; and

r sampling can be done at the Nyquist rate of the IF rather than the Nyquist
rate of the transmission.
The downconversion to an intermediate frequency (followed by bandpass ltering
to extract the passband around the IF) can be accomplished in two ways: by a
local oscillator modulating from above the carrier frequency (called high-side
injection) or from below (low-side injection). To see this, consider the double
sideband modulation (from Section 5.2) that creates the transmitted signal
v(t) = 2w(t)cos(2fc t)
from the message signal w(t) and the downconversion to IF via
x(t) = 2[v(t) + n(t)]cos(2fI t),
where n(t) represents interference such as noise and spurious signals from other
users. By virtue of the frequency-shifting property (A.33),
V (f ) = W (f + fc ) + W (f fc ),

(5.7)

and the spectrum of the IF signal is

X(f ) = V (f + fI ) + V (f fI ) + N (f + fI ) + N (f fI )
= W (f + fc fI ) + W (f fc fI ) + W (f + fc + fI )
+ W (f fc + fI ) + N (f + fI ) + N (f fI ).

(5.8)

94

Chapter 5. Analog (De)modulation

|V( f ) + N( f )|

1780

850

2175 1385 1245

3085 2156

105

850

1780

(a)

|X( f )|

|X( f )|

f (kHz)

105

f (kHz)

455 290

1245

1385

2175

1200 1410

2156

3085

500

f (kHz)

455
475

455

(b)

500

1410 1200

290

455
(c)

475

Figure 5.12 Example of high-side and low-side injection to an IF at fI = 455 kHz:

(a) transmitted spectrum, (b) low-side injected spectrum, and (c) high-side injected
spectrum.

Example 5.1. Consider a message spectrum W (f ) that has a bandwidth of

200 kHz, an upconversion carrier frequency fc = 850 kHz, and an objective to
downconvert to an intermediate frequency of fI = 455 kHz. For low-side injection, where the frequency of the local oscillator is f < fc , the goal is to center W (f fc + f ) in (5.8) at fI , so that f = fc 455 = 395. For high-side
injection (with f > fc ), the goal is to center W (f + fc f ) at fI , so that
f = fc + 455 = 1305. For illustrative purposes, Figure 5.12 supposes that the
interference N (f ) consists of a pair of delta functions at 105 and 1780 kHz.
Figure 5.12 sketches |V (f )| and |X(f )| for both high-side and low-side injection.
In this example, both methods end up with unwanted narrowband interferences
in the passband.
Observe the following.

r Low-side injection results in symmetry in the translated message spectrum

about fc on each of the positive and negative half-axes.

r High-side injection separates the undesired images further from the lowerr

frequency portion (which will ultimately be retained to reconstruct the message). This eases the requirements on the bandpass lter.
Both high-side and low-side injection can place frequency interferers in undesirable places. This highlights the need for adequate out-of-band rejection by
a bandpass lter before downconversion to IF.

95

Analog (De)modulation

|W( f )|
w(t)

22

22 f (kHz)

+
cos(2fct)

n(t)

r(t)

x(t)

cos(2frt)

Figure 5.13 Transmission system for Exercise 5.17.

Exercise 5.17. Consider the system described in Figure 5.13. The message w(t)
has a bandwidth of 22 kHz and a magnitude spectrum as shown. The message is
upconverted by a mixer with carrier frequency fc . The channel adds an interferer
n. The received signal r is downconverted to the IF signal x(t) by a mixer with
frequency fr .
a. With n(t) = 0, fr = 36 kHz, and fc = 83 kHz, indicate all frequency ranges
(i)(x) that include any part of the IF passband signal x(t).
(i) 020 kHz, (ii) 2040 kHz, (iii) 4060 kHz, (iv) 6080 kHz, (v) 80100 kHz,
(vi) 100120 kHz, (vii) 120140 kHz, (viii) 140160 kHz, (ix) 160180 kHz,
(x) 180200 kHz.
b. With fr = 36 kHz and fc = 83 kHz, indicate all frequency ranges (i)(x) that
include any frequency that causes a narrowband interferer n to appear in the
nonzero portions of the magnitude spectrum of the IF passband signal x(t).
c. With fr = 84 kHz and fc = 62 kHz, indicate every range (i)(x) that includes
any frequency that causes a narrowband interferer n to appear in the nonzero
portions of the magnitude spectrum of the IF passband signal x(t).
Exercise 5.18. A transmitter operates as a standard AM with suppressed carrier
transmitter (as in AM.m). Create a demodulation routine that operates in two
steps, by mixing with a cosine of frequency 3fc /4 and subsequently mixing with
a cosine of frequency fc /4. Where must pass/reject lters be placed in order to
ensure reconstruction of the message? Let fc = 2000.
Exercise 5.19. Consider the scheme shown in Figure 5.14. The absolute bandwidth of the baseband signal x1 is 4 kHz, f1 = 28 kHz, f2 = 20 kHz, and f3 = 26
kHz.
a.
b.
c.
d.
e.

What
What
What
What
What

is
is
is
is
is

the
the
the
the
the

absolute bandwidth of x2 (t)?

absolute bandwidth of x3 (t)?
absolute bandwidth of x4 (t)?
maximum frequency in x2 (t)?
maximum frequency in x3 (t)?

Exercise 5.20. Using your Matlab code from Exercise 5.18, investigate the
eect of a sinusoidal interference

96

Chapter 5. Analog (De)modulation

x1(t)

x2(t)

X2

BPF

Figure 5.14 Transmission system

lower cutoff f2
upper cutoff f3

Squaring
nonlinearity

cos(2f1t)

x4(t)

x3(t)

for Exercise 5.19.

cos(2f3t) + cos(2f4t)

|X1( f )|
1

x1(t)

f1

f1
x4(t)

x5(kTs)

x2(t)

x3(t)

lower cutoff f5
upper cutoff f6

cos(2f2t)
x6(kTs)

x7(kTs)
LPF

Ts=1/f7

x4(t)
BPF

x8(kT)
M

cutoff f9
cos(2f8t)

Figure 5.15 Transmission system for Exercise 5.21.

a. at frequency fc /6,
b. at frequency fc /3,
c. at frequency 3fc .
Exercise 5.21. Consider the PAM communication system in Figure 5.15. The
input x1 (t) has a triangular baseband magnitude spectrum. The frequency specications are f1 = 100 kHz, f2 = 1720 kHz, f3 = 1940 kHz, f4 = 1580 kHz,
f5 = 1720 kHz, f6 = 1880 kHz, and f7 = 1300 kHz.
a. Draw the magnitude spectrum |X5 (f )| between 3000 kHz. Be certain to
give specic values of frequency and magnitude at all breakpoints and local
maxima.
b. Specify values of f8 and f9 for which the system can recover the original
message without corruption with M = 2.

Exercise 5.22. This problem asks you to build a receiver from a limited number
of components. The parts available are
a. two product modulators with input u and output y related by
y(t) = u(t) cos(2fc t)
and carrier frequencies fc of 12 MHz and 50 MHz,
b. two linear bandpass lters with ideal rectangular magnitude spectrum of unity
gain between fU and fL and between fL and fU and zero elsewhere with
(fL , fU ) of (12 MHz, 32 MHz) and (35 MHz, 50 MHz),

Analog (De)modulation

97

|R( f )|
(1)

(1)
Figure 5.16 Spectrum of
the received signal for
Exercise 5.22.

46 40 36 30

10

10

30 36 40

46

f (MHz)

c. two impulse samplers with input u and output y related by

y(t) =

u(t)(t kTs )

k=

with sample periods of 1/15 and 1/12 s,

d. one square-law device with input u and output y related by
y(t) = u2 (t),
e. three summers with inputs u1 and u2 and output y related by
y(t) = u1 (t) + u2 (t).
The spectrum of the received signal is illustrated in Figure 5.16. The desired
baseband output of the receiver should be a scaled version of the triangular
portion centered at zero frequency with no other signals in the range between
8 and 8 MHz. Using no more than four parts from the ten available, build
a receiver that produces the desired baseband signal. Draw its block diagram.
Sketch the magnitude spectrum of the output of each part in the receiver.

For Further Reading

A friendly and readable introduction to analog transmission systems can be found
in

r P. J. Nahin, On the Science of Radio, AIP Press, 1996.

Sampling with Automatic Gain

Control

As foreshadowed in Section 2.8, transmission systems cannot be fully digital

because the medium through which the signal propagates is analog. Hence,
whether the signal begins as analog (such as voice or music) or whether it begins
as digital (such as mpeg, jpeg or wav les), it will be converted into a highfrequency analog signal when it is transmitted. In a digital receiver, the received
signal must be transformed into a discrete-time signal in order to allow subsequent digital processing.
This chapter begins by considering the sampling process both in the time
domain and in the frequency domain. Then Section 6.3 discusses how Matlab
can be used to simulate the sampling process. This is not completely obvious
because analog signals cannot be represented exactly in the computer. Two simple tricks are suggested. The rst expresses the analog signal in functional form
and takes samples at the desired times. The second oversamples the analog signal
so that it is represented at a high data rate; the sampling can then be done
on the oversampled signal.
Sampling and quantization are important because they translate the signal
from analog to digital. It is equally important to be able to translate from digital
back into analog, and the celebrated Nyquist sampling theorem shows that this is
possible for any bandlimited signal, assuming the sampling rate is fast enough.
When the goal of this translation is to rebuild a copy of the transmitted signal,
this is called reconstruction. When the goal is to determine the value of the signal
at some particular point, it is called interpolation. Techniques (and Matlab
code) for both reconstruction and interpolation appear in Section 6.4.
Figure 6.1 shows the received signal passing through a BPF (which removes
out-of-band interference and isolates the desired frequency range) followed by a
xed demodulation to the intermediate frequency (IF) at which sampling takes
place. The automatic gain control (AGC) accounts for changes in the strength
of the received signal. When the received signal is powerful, the gain a is small;
when the signal strength is low, the gain a is high. The goal is to guarantee that
the analog-to-digital converter does not saturate (the signal does not routinely

Sampling with Automatic Gain Control

Antenna

BPF

Analog
r(t)
conversion
to IF

Analog
received
signal
AGC

Sampler
& Quantizer
s(kTs) = s[k]

Quality
Assessment

99

Figure 6.1 The front end of

the receiver. After ltering

and demodulation, the
signal is sampled. An
automatic gain control
(AGC) is needed to utilize
the full dynamic range of
the quantizer.

surpass the highest level that can be represented), and that it does not lose
dynamic range (the digitized signal does not always remain in a small number
of the possible levels). The key in the AGC is that the gain must automatically
adjust to account for the signal strength, which may vary slowly over time.
The AGC provides the simplest example of a system element that must adapt
to changes in its environment (recall the fth element of Chapter 3). How
can such elements be designed? Software Receiver Design suggests a general
method based on gradient-directed optimization. First, a goal and an associated objective function are chosen. Since it is desired to maintain the output
of the AGC at a roughly constant power, the associated objective function is
dened to be the average deviation of the power from that constant; the goal is
to minimize the objective function. The gain parameter is then adjusted according to a steepest descent method that moves the estimate downhill towards
the optimal value that minimizes the objective. In this case the adaptive gain
parameter is increased (when the average power is too small) or decreased (when
the average power is too large), thus maintaining a steady power. While it would
undoubtedly be possible to design a successful AGC without recourse to such a
general optimization method, the framework developed in Sections 6.5 through
6.7 will also be useful in designing other adaptive elements such as the phasetracking loops of Chapter 10, the clock recovery algorithms of Chapter 12, and
the equalization schemes of Chapter 13.

6.1

Sampling and Aliasing

Sampling can be modeled as a point-by-point multiplication in the time domain
by a pulse train (a sequence of impulses). (Recall Figure 3.8 on page 49.) While
this is intuitively plausible, it is not terribly insightful. The eects of sampling
become apparent when viewed in the frequency domain. When the sampling
is done correctly, no information is lost. However, if the sampling is done too
slowly, aliasing artifacts are inevitable. This section shows the how and why
of sampling.

100

Chapter 6. Sampling with Automatic Gain Control

Signal w(t)

Pulse train
(t kTs)

Figure 6.2 An analog signal w(t)

is multiplied point-by-point by a
pulse train. This eectively
samples the analog signal at a
rate Ts .

Impulse
sampling
ws(t)
Point sampling
w[k] = w(kTs)
= w(t)|t = kTs

Suppose an analog waveform w(t) is to be sampled every Ts seconds to yield

a discrete-time sequence w[k] = w(kTs ) = w(t)|t=kTs for all integers k.1 This is
called point sampling because it picks o the value of the function w(t) at the
points kTs . One way to model point sampling is to create a continuous-valued
function that consists of a train of pulses that are scaled by the values w(kTs ).
The impulse sampling function is

ws (t) = w(t)

(t kTs ) =

k=

w(t)(t kTs )

k=

w(kTs )(t kTs ),

(6.1)

k=

and it is illustrated in Figure 6.2. The eect of multiplication by the pulse train
is clear in the time domain. But the relationship between ws (t) and w(t) is
clearer in the frequency domain, which can be understood by writing Ws (f ) as
a function of W (f ).
The transform Ws (f ) is given in (A.27) and (A.28). With fs = 1/Ts , this is
Ws (f ) = fs

W (f nfs ).

(6.2)

n=

Thus, the spectrum of the sampled signal ws (t) diers from the spectrum of the
original w(t) in two ways:

r Amplitude scalingeach term in the spectrum Ws (f ) is multiplied by the

factor fs .

r Replicasfor each n, Ws (f ) contains a copy of W (f ) shifted to f nfs .

Sampling creates an innite sequence of replicas, each separated by fs Hz. Said
another way, the process of sampling in time creates a periodicity in frequency,
1

Observe that the notation w(kTs ) means w(t) evaluated at the time t = kTs . This is also
notated w[k] (with the square brackets), where the sampling rate Ts is implicit.

Sampling with Automatic Gain Control

101

|W( f )|

fs B 0 B fs
2
2
|Ws( f )|

Figure 6.3 The spectrum

of a sampled signal is
periodic with period
equal to fs . In case (a),
the original spectrum
W (f ) is bandlimited to
less than fs /2 and there
is no overlapping of the
replicas. When W (f ) is
not bandlimited to less
than fs /2, as in (b), the
overlap of the replicas is
called aliasing.

2fs

fs
B 0 B
fs
fs B
fs+B
(a)
|W( f )|

B fs
2

fs
2

|Ws( f )|

fs

B
B
0
fs
fs+B
fs B

(b)

where the period is dened by the sampling rate. Readers familiar with Fourier
series will recognize this as the dual of the property that periodic in time is
the equivalent of sampling in frequency. Indeed, Equation (6.2) shows why the
relationships in Figure 3.9 on page 50 hold.
Figure 6.3 shows these replicas in two possible cases. In (a), fs 2B, where B
is the bandwidth of w(t), and the replicas do not overlap. Hence, it is possible to
extract the one replica centered at zero by using a lowpass lter. Assuming that
the ltering is without error, W (f ) is recovered from the sampled version Ws (f ).
Since the transform is invertible, this means that w(t) can be recovered from
ws (t). Therefore, no loss of information occurs in the sampling process.2 This
result is known as the Nyquist sampling theorem, and the minimum allowable
sampling rate is called the Nyquist rate.
2

Be clear about this: the analog signal w(t) is sampled to give ws (t), which is nonzero only
at the sampling instants kTs . If ws (t) is then input into a perfect analog lowpass lter, its
output is the same as the original w(t). Such ltering cannot be done with any digital lter
operating at the sampling rate fs . In terms of Figure 6.3, the digital lter can remove and
reshape the frequencies between the bumps, but can never remove the periodic bumps.

102

Chapter 6. Sampling with Automatic Gain Control

Nyquist Sampling Theorem: If the signal w(t) is bandlimited

to B (W (f ) = 0 for all |f | > B) and the sampling rate is
faster than fs = 2B, then w(t) can be reconstructed exactly
from its samples w(kTs ).
On the other hand, in part (b) of Figure 6.3, the replicas overlap because the
repetitions are narrower than the width of the spectrum W (f ). In this case, it is
impossible to recover the original spectrum perfectly from the sampled spectrum,
and hence it is impossible to exactly recover the original waveform from the
sampled version. The overlapping of the replicas and the resulting distortions in
the reconstructed waveform are called aliasing.
Bandwidth can also be thought of as limiting the rate at which data can
ow over a channel. When a channel is constrained to a bandwidth 2B, then the
output of the channel is a signal with bandwidth no greater than 2B. Accordingly,
the output can contain no frequencies above fs , and symbols can be transmitted
no faster than one every Ts seconds, where 1/Ts = fs .
Exercise 6.1. Human hearing extends up to about 20 kHz. What is the minimum sampling rate needed to fully capture a musical performance? Compare
this with the CD sampling rate of 44.1 kHz. Some animal sounds, such as the
singing of dolphins and the chirping of bats, occur at frequencies up to about 50
kHz. What does this imply about CD recordings of dolphin or bat sounds?
Exercise 6.2. US high-denition (digital) television (HDTV) is transmitted in
the same frequency bands as conventional television (for instance, Channel 2 is
at 54 MHz), and each channel has a bandwidth of about 6 MHz. What is the
minimum sampling rate needed to fully capture the HDTV signal once it has
been demodulated to baseband?
Exercise 6.3. An analog signal has nonzero values in its magnitude spectrum at
every frequency between B and B. This signal is sampled with period T where
1/T > B. TRUE or FALSE: The discrete-time signal can have components in its
spectrum at frequencies between B and 1/T .
Exercise 6.4. The triangularly shaped magnitude spectrum of a real message
signal w(t) is shown in Figure 6.4, where B = 0.2 MHz. The received signal
r(t) = 0.15w(t) cos(2f t)
is modulated by an AM with suppressed carrier with f = 1.45 MHz, and attentuated. With 1/f < T1 < 1/(2B), select T1 , T2 , T3 , and so that the magnitude
spectrum of x3 [k] matches the magnitude spectrum of T1 -spaced samples of w(t).
Justify your answer by drawing the magnitude spectra of x1 , x2 , and x3 between
f and f .

103

Sampling with Automatic Gain Control

|W( f )|

x1[k]

r(t)

x2[k]

x3[k]
LPF

kT1
cos(2kT2)

cutoff frequency: 1/T3

passband gain: 1
stopband gain: 0

Figure 6.4 The triangularly shaped magnitude spectrum of the real message signal

w(t) is shown on the left. The receiver structure in Exercise 6.4 is shown on the right.
|X( f )|

v1(t)

x(t)

v2(kTs)

v3(kTs)

X2
square law
450 400 350

350 400 450 f

v4(kTs)
LPF

Ts = 0.001 s
cos(400kTs)

Figure 6.5 The signal x(t) with spectrum X(f ) is input into this communication

system. Exercise 6.5 asks for the absolute bandwidth of the signal at each point as it
moves through the system.

Exercise 6.5. The signal x(t) with magnitude spectrum shown in the left part
of Figure 6.5 is input to the transmission system on the right. The lowpass lter
with output v4 (kTs ) has a cuto frequency of 200 Hz, a passband gain of 1,
and a stopband gain of zero. Specify all frequencies between zero and 2000 Hz
at which the magnitude spectra of |V1 (f )|, |V2 (f )|, |V3 (f )|, and |V4 (f )| have
nonzero values.

6.2

Downconversion via Sampling

The processes of modulation and demodulation, which shift the frequencies of
a signal, can be accomplished by mixing with a cosine wave that has a frequency equal to the amount of the desired shift, as was demonstrated repeatedly
throughout Chapter 5. But this is not the only way. Since sampling creates a
collection of replicas of the spectrum of a waveform, it changes the frequencies
of the signal.
When the message signal is analog and bandlimited to B, sampling can be
used as a step in the demodulation process. Suppose that the signal is transmitted
with a carrier at frequency fc . Direct sampling of this signal creates a collection
of replicas, one near DC. This procedure is shown in Figure 6.6 for fs = fc /2,
though beware: when fs and fc are not simply related, the replica might not
land exactly at DC.
This demodulation by sampling is diagrammed in Figure 6.7 (with fs = fc /n,
where n is a small positive integer), and can be thought of as an alternative to
mixing with a cosine (that must be synchronized in frequency and phase with
the transmitter oscillator). The magnitude spectrum |W (f )| of a message w(t)

104

Chapter 6. Sampling with Automatic Gain Control

|W( f )|
1

Figure 6.6 Spectra in a

(a)

|S( f )|
1/2

fcB fc+B
fc

(b)

fcB fc+B
fc

|Y( f )|

3fc /2

fc

fc /2

0
(c)

fc /2
(= fc fs)

fc

sampling downconverter
such as Figure 6.7. The
(bandlimited analog) signal
W (f ) shown in (a) is
upconverted to the
transmitted signal in (b).
Directly sampling this (at a
rate equal to fs = fc /2)
results in the spectrum
shown in the bottom plot.

3fc /2
f
(= fc+fs)

is shown in Figure 6.6(a), and the spectrum after upconversion is shown in part
(b); this is the transmitted signal s(t). At the receiver, s(t) is sampled, which
can be modeled as a multiplication with a train of delta functions in time,
y(t) = s(t)

(t nTs ),

n=

where Ts is the sample period.

Using (6.2), this can be transformed into the frequency domain as

1 
Y (f ) =
S(f nfs ),
Ts n=

where fs = 1/Ts . The magnitude spectrum of Y (f ) is illustrated in Figure 6.6(c)

for the particular choice fs = fc /2 (and Ts = 2/fc ) with B < fc /4 = fs /2.
There are three ways in which the sampling can proceed:
1. sample faster than the Nyquist rate of the IF frequency,
2. sample slower than the Nyquist rate of the IF frequency, and then downconvert
the replica closest to DC, and
3. sample so that one of the replicas is directly centered at DC.
The rst is a direct imitation of the analog situation where no aliasing will occur.
This may be expensive because of the high sample rates required to achieve
Nyquist sampling. The third is the situation depicted in Figures 6.6 and 6.7,
which permit downconversion to baseband without an additional oscillator. This
may be sensitive to small deviations in frequency (for instance, when fs is not

Sampling with Automatic Gain Control

w(t)

s(t)

105

m(t)

y(t)
LPF
Ts

fcutoff = B

cos(2fct)
Figure 6.7 System diagram showing how sampling can be used to downconvert a

signal. The spectra corresponding to w(t), s(t), and y(t) are shown in Figure 6.6. The
output of the LPF contains only the M-shaped portion nearest zero.

exactly fc /2). The middle method downconverts part of the way by sampling
and part of the way by mixing with a cosine. The middle method is used in the
M6 receiver project in Chapter 15.
Exercise 6.6. Create a simulation of a sampling-based modulator that takes a
signal with bandwidth 100 Hz and transforms it into the same signal centered
at 5000 Hz. Be careful; there are two sampling rates in this problem. One
reects the assumed sampling rate for the modulation and the other represents
the sampling rate that is used in Matlab to represent a continuous time
signal. You may wish to reuse code from sine100hzsamp.m. What choices have
you made for these two sampling rates?
Exercise 6.7. Implement the procedure diagrammed in Figure 6.7. Comment
on the choice of sampling rates. How have you specied the LPF?
Exercise 6.8. Using your code from Exercise 6.7, examine the eect of incorrect sampling rates by demodulating with fs + instead of fs . This is analogous
to the problem that occurs in cosine mixing demodulation when the frequency is
not accurate. Is there an analogy to the phase problem that occurs, for instance,
with nonzero in (5.4)?
Exercise 6.9. Consider the part of a communication system shown in Figure
6.8.
a. Sketch the magnitude spectrum |X1 (f )| of
x1 (t) = w(t)cos(1500t).
Be certain to give specic values of frequency and magnitude at all signicant
points in the sketch.
b. Draw the magnitude spectrum |X2 (f )| of
x2 (t) = w(t)x1 (t).
Be certain to give specic values of frequency and magnitude at all signicant
points in the sketch.

106

Chapter 6. Sampling with Automatic Gain Control

|W( f )|
x1(t)

w(t)

0.2

x2(t)

x3(t)
Ts

220

220

cos(1500t)

Figure 6.8 Input spectrum and system diagram for Exercise 6.9.

w(t)

x2[k]

x1(t)
kTs

cos(2fct) cos(2fIt)

x3[k]

LPF

x4[k]

cutoff frequency: f2
cos(2f1kTs)

Figure 6.9 The digital

receiver used in
Exercise 6.10.

c. Between 3750 Hz and 3750 Hz, draw the magnitude spectrum |X3 (f )| of
x3 (t) = x2 (t)

(t kTs ).

k=

for Ts = 400 s. Be certain to give specic values of frequency and magnitude

at all signicant points in the sketch.

Exercise 6.10. Consider the digital receiver shown in Figure 6.9. The baseband
signal w(t) has absolute bandwidth B and the carrier frequency is fc . The channel
adds a narrowband interferer at frequency fI . The received signal is sampled with
period Ts . As shown, the sampled signal is demodulated by mixing with a cosine
of frequency f1 and the ideal lowpass lter has a cuto frequency of f2 . For the
following designs you are to decide whether they are successful, i.e., whether or
not the magnitude spectrum of the lowpass lter output x4 is the same (up to
a scale factor) as the magnitude spectrum of the sampled w(t) with a sample
period of Ts .
a. Candidate System A: B = 7 kHz, fc = 34 kHz, fI = 49 kHz, Ts = 1/34 ms,
f1 = 0, and f2 = 16 kHz.
b. Candidate System B: B = 11 kHz, fc = 39 kHz, fI = 130 kHz, Ts = 1/52 ms,
f1 = 13 kHz, and f2 = 12 kHz.

Exercise 6.11. Consider the communication system shown in Figure 6.10. In

this problem you are to build a receiver from a limited number of components.
The parts available are

r four mixers with input u and output y related by

y(t) = u(t) cos(2fo t)
and oscillator frequencies fo of 1 MHz, 1.5 MHz, 2 MHz, and 4 MHz,

107

Sampling with Automatic Gain Control

|R( f )|
(1.5)
(1.5)

r(t)

y(t)
BPF
Ts

4 2

cos(2fot)

f (MHz)

Figure 6.10 The digital receiver used in Exercise 6.11.

r four ideal linear bandpass lters with passband (fL , fU ) of (0.5 MHz, 6 MHz),
(1.2 MHz, 6.2 MHz), (3.4 MHz, 7.2 MHz), and (4.2 MHz, 8.3 MHz),

r four impulse samplers with input u and output y related by

y(t) =

u(t)(t kTs )

k=

with sample periods of 1/7, 1/5, 1/4, and 1/3.5 microseconds.

The magnitude spectrum |R(f )| of the received signal r(t) is shown on the left in
Figure 6.10. The objective is to specify the bandpass lter, sampler, and mixer
so that the M-shaped segment of the magnitude spectrum is centered at f = 0
in the output |Y (f )| with no other signals within 1.5 MHz of the upper and
lower edges.
a. Specify the three parts from the 12 provided:
(i) bandpass lter passband range (fL , fU ) in MHz,
(ii) sampler period Ts in s,
(iii) mixer oscillator frequency fo in MHz.
b. For the three components selected in part (a), sketch the magnitude spectrum
of the sampler output between 20 and +20 MHz. Be certain to give specic
values of frequency and magnitude at all signicant points in the spectra.
c. For the three components selected in part (a), sketch the magnitude spectrum
of y(t) between between the frequencies 12 and +12 MHz for your design.
Be certain to give specic values of frequency and magnitude at all signicant
points in the spectra.
d. Is the magnitude spectrum of y(t) identical to the M-shaped segment of
|R(f )| rst downconverted to baseband and then sampled?

Exercise 6.12. The message signal u(t) and additive interferer w(t) with magnitude spectra shown in Figure 6.11(a) are applied to the system in Figure 6.11(b).
The analog mixer frequencies are fc = 1600 kHz and fd = 1240 kHz. The BPF
with output x(t) is assumed ideal, is centered at fc , and has lower cuto frequency
fL , upper cuto frequency fU , and zero phase at fc . The period of the sampler
1
104 s . The phase of the discrete-time mixer is assumed to be
is Ts = 71
adjusted to the value that maximizes the ratio of signal to interferer noise power
in y(kTs ). The LPF with output y(kTs ) is assumed ideal with cuto frequency

108

Chapter 6. Sampling with Automatic Gain Control

|W( f )|

|U( f )|

40 f (kHz) 40
(a)
w(t)

860 720

720 860

f (kHz)
(b)

x(t)

u(t)

y(kTs)

BPF

LPF

Figure 6.11 Magnitude spectra of

(a) the message and (b) the
interference signals considered in
Exercise 6.12. The receiver
structure is shown schematically in
(c).

Ts
cos(2fct)

x(t)

(c)

cos(2fdt) cos(2kTs +)

r(t)

y(kTs)
LPF
kTs

cos(2fct)

Figure 6.12 Digital receiver

considered in Exercise 6.13.

cos(2fckTs)

. The design objective is for the spectrum of y(kTs ) to estimate the spectrum
of a sampled u(t). You are to select the upper and lower cuto frequencies of the
BPF, the frequency of the discrete-time mixer, and the cuto frequency of the
LPF in order to meet this objective.
a. Design the desired BPF by specifying its upper fU and lower fL cuto frequencies.
b. Compute the desired discrete-time mixer frequency .
c. Design the desired LPF by specifying its cuto frequency .
Exercise 6.13. Consider the digital receiver in Figure 6.12 producing y(kTs ),
which is intended to match the input x(t) sampled every Ts seconds. The absolute bandwidth of x(t) is B. The carrier frequency fc is 10 times B. The sample
frequency 1/Ts is 2.5 times fc . Note that the sample frequency 1/Ts is above the
Nyquist frequency of the received signal r(t). Determine the maximum cuto
frequency as a function of the input bandwidth B for the lowpass lter producing y(kTs) so the design objective of matching samples of x(t) with a sample
frequency of 1/Ts is achieved.

6.3

Exploring Sampling in MATLAB

It is not possible to capture all of the complexities of analog-to-digital conversion
inside a computer program, because all signals within a (digital) computer are
already sampled. Nonetheless, most of the key ideas can be illustrated by using
two tricks to simulate the sampling process:

Sampling with Automatic Gain Control

109

r evaluate a function at appropriate values (or times),

r represent a data waveform by a large number of samples and then reduce the
number of samples.
The rst is useful when the signal can be described by a known function, while
the second is necessary whenever the procedure is data-driven, that is, when no
functional form is available. This section explores both approaches via a series
of Matlab experiments.
Consider representing a sine wave of frequency f = 100 Hz. The sampling
theorem asserts that the sampling rate must be greater than the Nyquist rate
of 200 samples per second. But, in order to visualize the wave clearly, it is
often useful to sample considerably faster. The following Matlab code calculates
and plots the rst 1/10 second of a 100 Hz sine wave with a sampling rate of
fs = 1/Ts = 10000 samples per second.
Listing 6.1. sine100hz.m generate 100 Hz sine wave with sampling rate fs=1/Ts
f =100;
time = 0 . 1 ;
Ts =1/10000;
t=Ts : Ts : time ;
w=s i n (2 * pi * f * t ) ;
plot ( t ,w)
xlabel ( seconds )
ylabel ( amplitude )

%
%
%
%
%
%

f r e q u e n c y o f wave
t o t a l time i n s e c o n d s
sampling i n t e r v a l
d e f i n e a time v e c t o r
d e f i n e th e s i n e wave
p l o t th e s i n e v s . time

Running sine100hz.m plots the rst 10 periods of the sine wave. Each period
lasts 0.01 s, and each period contains 100 points, as can be veried by looking
at w(1:100). Changing the variable time or Ts displays dierent numbers of
cycles of the same sine wave, while changing f plots sine waves with dierent
underlying frequencies.
Exercise 6.14. What must the sampling rate be so that each period of the wave
is represented by 20 samples? Check your answer using the program above.
Exercise 6.15. Let Ts=1/500. How does the plot of the sine wave appear? Let
Ts=1/100, and answer the same question. How large can Ts be if the plot is to
retain the appearance of a sine wave? Compare your answer with the theoretical
limit. Why are they dierent?
When the sampling is rapid compared with the underlying frequency of the signal (for instance, the program sine100hz.m creates 100 samples in each period),
then the plot appears and acts much like an analog signal, even though it is still,
in reality, a discrete-time sequence. Such a sequence is called oversampled relative
to the signal period. The following program simulates the process of sampling
the 100 Hz oversampled sine wave. This is downsampling, as shown in Figure
3.10 on page 51.

110

Chapter 6. Sampling with Automatic Gain Control

Amplitude

Figure 6.13 Removing all but one

of each N points from an
oversampled waveform simulates
the sampling process.

0.01

0.02
Seconds

0.03

0.04

Listing 6.2. sine100hzsamp.m simulated sampling of the 100 Hz sine wave

f =100; time = 0 . 0 5 ;
Ts =1/10000; t=Ts : Ts : time ;
w=s i n (2 * pi * f * t ) ;
s s =10;
wk=w( 1 : s s : end ) ;
ws=zeros ( s i z e (w ) ) ; ws ( 1 : s s : end)=wk ;
plot ( t , w)
hold on , plot ( t , ws , r ) , hold o f f

%
%
%
%
%
%
%

s e t time v e c t o r s
c r e a t e s i n e wave w( t )
t a k e 1 i n s s s a m p l es
th e sampled s e q u e n c e
sampled waveform ws ( t )
p l o t th e waveform
p l o t sampled wave

Running sine100hzsamp.m results in the plot shown in Figure 6.13, where the
continuous sine wave w is downsampled by a factor of ss=10; that is, all but
one of each ss samples is removed. Thus, the waveform w represents the analog
signal that is to be sampled at the eective sampling interval ss*Ts. The spiky
signal ws corresponds to the sampled signal ws (t), while the sequence wk contains
just the amplitude values at the tips of the spikes.
Exercise 6.16. Modify sine100hzsamp.m to create an oversampled sinc wave,
and then sample this with ss=10. Repeat this exercise with ss=30, ss=100, and
ss=200. Comment on what is happening. Hint: in each case, what is the eective
sampling interval?
Exercise 6.17. Plot the spectrum of the 100 Hz sine wave when it is created
with dierent downsampling rates ss=10, ss=11, ss=30, and ss=200. Explain
what you see.

6.4

Interpolation and Reconstruction

The previous sections explored how to convert analog signals into digital signals.
The central result is that, if the sampling is done faster than the Nyquist rate,
then no information is lost. In other words, the complete analog signal w(t)
can be recovered from its discrete samples w[k]. When the goal is to nd the
complete waveform, this is called reconstruction; when the goal is to nd values
of the waveform at particular points between the sampling instants, it is called

Sampling with Automatic Gain Control

111

interpolation. This section explores bandlimited interpolation and reconstruction

in theory and practice.
The samples w(kTs ) form a sequence of numbers that represent an underlying
continuous-valued function w(t) at the time instants t = kTs . The sampling interval Ts is presumed to have been chosen so that the sampling rate fs > 2B, where
B is the highest frequency present in w(t). The Nyquist sampling theorem presented in Section 6.1 states that the values of w(t) can be recovered exactly at
any time . The formula (which is justied subsequently) for recovering w( )
from the samples w(kTs ) is

w( ) =
ws (t) sinc ( t) dt,
t=

where ws (t) (dened in (6.1)) is zero everywhere except at the sampling instants
t = kTs . Substituting (6.1) into w( ) shows that this integral is identical to the
sum




kTs
.
(6.3)
w( ) =
w(kTs ) sinc
Ts
k=

In principle, if the sum is taken over all time, the value of w( ) is exact. As a
practical matter, the sum must be taken over a suitable (nite) time window.
To see why interpolation works, note that the formula (6.3) is a convolution
(in time) of the signal w(kTs ) and the sinc function. Since convolution in time
is the same as multiplication in frequency by (A.40), the transform of w( ) is
equal to the product of F {ws (kTs )} and the transform of the sinc. By (A.22),
the transform of the sinc function in time is a rect function in frequency. This
rect function is a lowpass lter, since it passes all frequencies below fs /2 and
removes all frequencies above. Since the process of sampling a continuous-time
signal generates replicas of the spectrum at integer multiples of fs by (6.2), the
lowpass lter removes all but one of these replicas. In eect, the sampled data
are passed through an analog lowpass lter to create a continuous-time function,
and the value of this function at time is the required interpolated value. When
= nTs , then sinc( nTs ) = 1, and sinc( nTs ) = 0 for all kTs with k = n.
When is between sampling instants, the sinc is nonzero at all kTs , and (6.3)
combines them to recover w( ).
To see how (6.3) works, the following code generates a sine wave w of frequency
20 Hz with a sampling rate of 100 Hz. This is a modestly sampled sine wave,
having only ve samples per period, and its graph is jumpy and discontinuous.
Because the sampling rate is greater than the Nyquist rate, it is possible in
principle to recover the underlying smooth sine wave from which the samples
are drawn. Running sininterp.m shows that it is also possible in practice. The
plot in Figure 6.14 shows the original wave (which appears choppy because it is
sampled only ve times per period), and the reconstructed or smoothed waveform
(which looks just like a sine wave). The variable intfac species how many extra

Chapter 6. Sampling with Automatic Gain Control

1
Amplitude

112

Figure 6.14 A convincing sine wave

1
10

10.1

10.2
Time

10.3

10.4

can be reconstructed from its

samples using sinc interpolation.
The choppy wave represents the
samples, and the smooth wave
shows the reconstruction.

interpolated points are calculated, and need not be an integer. Larger numbers
result in smoother curves but also require more computation.
Listing 6.3. sininterp.m demonstrate interpolation/reconstruction using sin wave
f =20; Ts =1/100; time =20;
% s a m p l i n g i n t e r v a l and time
t=Ts : Ts : time ;
% time v e c t o r
w=s i n (2 * pi * f * t ) ;
% w( t ) = a s i n e wave o f f H e r tz
o v e r =100;
% # o f data p o i n t s i n smoothing
i n t f a c =10;
% how many i n t e r p o l a t e d p o i n t s
tnow =10.0/ Ts : 1 / i n t f a c : 1 0 . 5 / Ts ;
% i n t e r p o l a t e from 10 to 1 0 . 5
wsmooth=zeros ( s i z e ( tnow ) ) ;
% s a v e smoothed data h e r e
f o r i =1: length ( tnow )
% and l o o p f o r n e x t p o i n t
wsmooth ( i )= i n t e r p s i n c (w, tnow ( i ) , o v e r ) ;
end

In implementing (6.3), some approximations are used. First, the sum cannot
be calculated over an innite time horizon, and the variable over replaces the
over

sum
k= with
k=over . Each pass through the for loop calculates one
point of the smoothed curve wsmooth using the Matlab function interpsinc.m,
which is shown below. The value of the sinc is calculated at each time using the
function srrc.m with the appropriate oset tau, and then the convolution is
performed by the conv command. This code is slow and unoptimized. A clever
programmer will see that there is no need to calculate the sinc for every point, and
ecient implementations use sophisticated look-up tables to avoid the calculation
of transcendental functions completely.
Listing 6.4. interpsinc.m interpolate to nd a single point using the direct method
function y=i n t e r p s i n c ( x , t , l , beta )
% x = sampled data
% t = p l a c e a t which v a l u e d e s i r e d
% l = ones i d e d l e n g t h o f data to i n t e r p o l a t e
% b eta = r o l l o f f f a c t o r f o r s r r c f u n c t i o n
%
= 0 is a sinc
i f nargin==3, beta =0; end ;
% i f u n s p e c i f i e d , b e ta i s 0
tnow=round ( t ) ;
% s e t i n d e x tnow=i n t e g e r p a r t
tau=tround ( t ) ;
% p l u s tau=f r a c t i o n a l p a r t
% i n t e r p o l a t e s i n c a t o f f s e t tau
s t a u=s r r c ( l , beta , 1 , tau ) ;
x t a u=conv ( x ( tnowl : tnow+l ) , s t a u ) ; % i n t e r p o l a t e th e s i g n a l
% to g e t th e new sample
y=x t a u (2 * l +1);

While the indexing needed in interpsinc.m is a bit tricky, the basic idea is
not: the sinc interpolation of (6.3) is just a linear lter with impulse response
h(t) = sinc(t). (Remember, convolutions are the hallmark of linear lters.) Thus,

Sampling with Automatic Gain Control

113

it is a lowpass lter, since the frequency response is a rect function. The delay
is proportional to the phase of the frequency response.
Exercise 6.18. In sininterp.m, what happens when the sampling rate is too
low? How large can the sampling interval Ts be? How high can the frequency f
be?
Exercise 6.19. In sininterp.m, what happens when the window is reduced?
Make over smaller and nd out. What happens when too few points are interpolated? Make intfac smaller and nd out.
Exercise 6.20. Create a more interesting (more complex) wave w(t). Answer
the above questions for this w(t).
Exercise 6.21. Let w(t) be a sum of ve sinusoids for t between 10 and 10 s.
Let w(kT ) represent samples of w(t) with T = 0.01 s. Use interpsinc.m to interpolate the values w(0.011), w(0.013), and w(0.015). Compare the interpolated
values with the actual values. Explain any discrepancies.
Observe that sinc(t) dies away (slowly) in time at a rate proportional to
1/(t). This is one of the reasons why so many terms are used in the convolution (i.e., why the variable over is large). A simple way to reduce the number of terms is to use a function that dies away more quickly than the sinc; a
common choice is the square-root raised cosine (SRRC) function, which plays
an important role in pulse shaping in Chapter 11. The functional form of the
SRRC is given in Equation (11.8). The SRRC can easily be incorporated into
the interpolation code by replacing the code interpsinc(w,tnow(i),over) with
interpsinc(w,tnow(i),over,beta).
Exercise 6.22. With beta=0, the SRRC is exactly the sinc. Redo the above
exercises trying various values of beta between 0 and 1.
The function srrc.m is available on the website. Its help le is
%
%
%
%
%
%
%
%

s=s r r c ( syms , beta , P , t o f f ) ;

G ener ate a SquareRoot R a i s e d C o s i n e P u l s e
syms i s 1/2 th e l e n g t h o f s r r c p u l s e
i n symbol d u r a t i o n s
beta i s th e r o l l o f f f a c t o r :
b eta =0 g i v e s th e s i n c f u n c t i o n
P i s th e o v e r s a m p l i n g f a c t o r
t o f f i s th e phase ( o r t i m i n g ) o f f s e t

Matlab also has a built-in function called resample, which has the following
help le:
% Change th e s a m p l i n g r a t e o f a s i g n a l
% Y = r e s a m p l e (X, P ,Q) r e s a m p l e s th e s e q u e n c e
% i n v e c t o r X a t P/Q t i m e s th e o r i g i n a l sample

114

Chapter 6. Sampling with Automatic Gain Control

% r a t e u s i n g a p o l y p h a s e i m p l em en ta ti o n .
% Y i s P/Q t i m e s th e l e n g t h o f X.
% P and Q must be p o s i t i v e i n t e g e r s .

This technique is dierent from that used in (6.3). It is more ecient numerically at reconstructing entire waveforms, but it works only when the desired
resampling rate is rationally related to the original. The method of (6.3) is far
more ecient when isolated (not necessarily evenly spaced) interpolating points
are required, which is crucial for synchronization tasks in Chapter 12.

6.5

Iteration and Optimization

An important practical part of the sampling procedure is that the dynamic range
of the signal at the input to the sampler must remain within bounds. This can be
accomplished using an automatic gain control, which is depicted in Figure 6.1 as
multiplication by a scalar a, along with a quality assessment block that adjusts
a in response to the power at the output of the sampler. This section discusses
the background needed to understand how the quality assessment works. The
essential idea is to state the goal of the assessment mechanism as an optimization
problem.
Many problems in communications (and throughout engineering) can be
framed in terms of an optimization problem. Solving such problems requires
three basic steps:
1. setting a goalchoosing a performance or objective function,
2. choosing a method of achieving the goalminimizing or maximizing the objective function, and
3. testing to make sure the method works as anticipated.
Setting the goal usually consists of nding a function that can be minimized
(or maximized), and for which locating the minimum (or maximum) value provides useful information about the problem at hand. Moreover, the function
must be chosen carefully so that it (and its derivative) can be calculated from
quantities that are known, or that can be derived from signals that are easily
obtainable. Sometimes the goal is obvious, and sometimes it is not.
There are many ways of carrying out the minimization or maximization procedure. Some of these are direct. For instance, if the problem is to nd the point
at which a polynomial function achieves its minimum value, this can be solved
directly by nding the derivative and setting it equal to zero. Often, however,
such direct solutions are impossible, and, even when they are possible, recursive
(or adaptive) approaches often have better properties when the signals are noisy.
This chapter focuses on a recursive method called steepest descent, which is the
basis of many adaptive elements used in communications systems (and of all the
elements used in Software Receiver Design).

115

Sampling with Automatic Gain Control

The nal step in implementing any solution is to check that the method
behaves as desired, despite any simplifying assumptions that may have been
made in its derivation. This may involve a detailed analysis of the resulting
methodology, or it may involve simulations. Thorough testing would involve both
analysis and simulation in a variety of settings that mimic, as closely as possible,
the situations in which the method will be used.
Imagine being lost on a mountainside on a foggy night. Your goal is to
get to the village which lies at the bottom of a valley below. Though
you cannot see far, you can reach out and feel the nearby ground. If you
repeatedly step in the direction that heads downhill most steeply, you
eventually reach a depression in which all directions lead up. If the contour of the land is smooth, and without any local depressions that can
trap you, then you will eventually arrive at the village. The optimization
procedure called steepest descent implements this scenario mathematically, where the mountainside is dened by the performance function
and the optimal answer lies in the valley at the minimum value. Many
standard communications algorithms (adaptive elements) can be viewed
in this way.

6.6

An Example of Optimization: Polynomial Minimization

This rst example is too simple to be of practical use, but it does show many
of the ideas starkly. Suppose that the goal is to nd the value at which the
polynomial
J(x) = x2 4x + 4

(6.4)

achieves its minimum value. Thus step (1) is set. As any calculus book will
suggest, the direct way to nd the minimum is to take the derivative, set it
equal to zero, and solve for x. Thus, dJ(x)/dx = 2x 4 = 0 is solved when x =
2. This is indeed the value of x for which the parabola J(x) reaches bottom.
Sometimes (one might truthfully say often), however, such direct approaches
are impossible. Maybe the derivative is just too complicated to solve (which can
happen when the functions involved in J(x) are extremely nonlinear). Or maybe
the derivative of J(x) cannot be calculated precisely from the available data, and
instead must be estimated from a noisy data stream.
One alternative to the direct solution technique is an adaptive method called
steepest descent (when the goal is to minimize) or hill climbing (when the
goal is to maximize). Steepest descent begins with an initial guess of the location of the minimum, evaluates which direction from this estimate is most steeply
downhill, and then makes a new estimate along the downhill direction. Similarly, hill climbing begins with an initial guess of the location of the maximum,
evaluates which direction climbs the most rapidly, and then makes a new esti-

116

Chapter 6. Sampling with Automatic Gain Control

J(x)
Gradient direction
has same sign as
the derivative
Derivative is
slope of line
tangent to
J(x) at x

xn

Minimum

Figure 6.15 Steepest descent

nds the minimum of a function

by always pointing in the
direction that leads downhill.

xp

Arrows point in minus gradient

direction towards the minimum

mate along the uphill direction. With luck, the new estimates are better than
the old. The process repeats, ideally getting closer to the optimal location at
each step. The key ingredient in this procedure is to recognize that the uphill
direction is dened by the gradient evaluated at the current location, while the
downhill direction is the negative of this gradient.
To apply steepest descent to the minimization of the polynomial J(x) in (6.4),
suppose that a current estimate of x is available at time k, which is denoted x[k].
A new estimate of x at time k + 1 can be made using

dJ(x) 
x[k + 1] = x[k]
,
(6.5)
dx x=x[k]
where is a small positive number called the stepsize, and the gradient (derivative) of J(x) is evaluated at the current point x[k]. This is then repeated again
and again as k increments. This procedure is shown in Figure 6.15. When the
current estimate x[k] is to the right of the minimum, the negative of the gradient
points left. When the current estimate is to the left of the minimum, the negative gradient points to the right. In either case, as long as the stepsize is suitably
small, the new estimate x[k + 1] is closer to the minimum than the old estimate
x[k]; that is, J(x[k + 1]) is less than J(x[k]).
To make this explicit, the iteration dened by (6.5) is
x[k + 1] = x[k] (2x[k] 4),
or, rearranging,
x[k + 1] = (1 2)x[k] + 4.

(6.6)

In principle, if (6.6) is iterated over and over, the sequence x[k] should approach
the minimum value x = 2. Does this actually happen?
There are two ways to answer this question. It is straightforward to simulate
the process. Here is some Matlab code that takes an initial estimate of x called
x(1) and iterates Equation (6.6) for N=500 steps.
Listing 6.5. polyconverge.m nd the minimum of J(x) = x2 4x + 4 via steepest
descent
N=500;

% number o f i t e r a t i o n s

Sampling with Automatic Gain Control

117

Starting values

4
3

Figure 6.16 The program

polyconverge.m attempts to
locate the smallest value of
J(x) = x2 4x + 4 by descending
the gradient. Fifty dierent
starting values all converge to the
same minimum at x = 2.

2
1
0

1
2
0

50

100

150

200 250
Time

mu= . 0 1 ;
x=zeros ( 1 ,N ) ;
x(1)=3;
f o r k =1:N1
x ( k+1)=(12*mu) * x ( k)+4 *mu ;
end

300

350

400

% algorithm s t e p s i z e
% i n i t i a l i z e x to z e r o
% s t a r t i n g point x (1)
% update e q u a t i o n

Figure 6.16 shows the output of polyconverge.m for 50 dierent x(1) starting
values superimposed; all converge smoothly to the minimum at x = 2.
Exercise 6.23. Explore the behavior of steepest descent by running
polyconverge.m with dierent parameters.
a. Try mu = 0.01, 0, 0.0001, 0.02, 0.03, 0.05, 1.0, 10.0. Can mu be too large or
too small?
b. Try N = 5, 40, 100, 5000. Can N be too large or too small?
c. Try a variety of values of x(1). Can x(1) be too large or too small?
As an alternative to simulation, observe that the process (6.6) is itself a linear
time-invariant system, of the general form
x[k + 1] = ax[k] + b,

(6.7)

which is stable as long as |a| < 1. For a constant input, the nal-value theorem of
z-transforms (see (A.55)) can be used to show that the asymptotic (convergent)
output value is limk xk = b/(1 a). To see this without reference to arcane
theory, observe that if xk is to converge, then it must converge to some value,
say x . At convergence, x[k + 1] = x[k] = x , so (6.7) implies that x = ax +
b, which implies that x = b/(1 a). (This holds assuming that |a| < 1.) For
example, for (6.6), x = 4/[1 (1 2)] = 2, which is indeed the minimum.
Thus, both simulation and analysis suggest that the iteration (6.6) is a viable
way to nd the minimum of the function J(x), as long as is suitably small. As
will become clearer in later sections, such solutions to optimization problems are
almost always possibleas long as the function J(x) is dierentiable. Similarly, it

118

Chapter 6. Sampling with Automatic Gain Control

is usually quite straightforward to simulate the algorithm to examine its behavior

in specic cases, though it is not always so easy to carry out a theoretical analysis.
By their nature, steepest-descent and hill-climbing methods use only local
information. This is because the update from a point x[k] depends only on the
value of x[k] and on the value of its derivative evaluated at that point. This can
be a problem, since if the objective function has many minima, the steepestdescent algorithm may become trapped at a minimum that is not (globally)
the smallest. These are called local minima. To see how this can happen, consider
the problem of nding the value of x that minimizes the function
J(x) = e0.1|x| sin(x).

(6.8)

Applying the chain rule, the derivative is

e0.1 |x| cos(x) 0.1e0.1 |x| sin(x) sign(x),
where


sign(x) =

1
1

x>0
x<0

(6.9)

is the formal derivative of |x|. Solving directly for the minimum point is nontrivial
(try it!). Yet implementing a steepest-descent search for the minimum can be
done in a straightforward manner using the iteration
x[k + 1] = x[k] e0.1 |x[k]| (cos(x[k]) 0.1 sin(x[k])sign(x)).

(6.10)

To be concrete, replace the update equation in polyconverge.m with

x ( k+1) = x ( k ) mu * exp ( 0.1 * abs ( x ( k ) ) ) * . . .
( cos ( x ( k ) ) 0 . 1 * s i n ( x ( k ) ) * sign ( x ( k ) ) ) ;

Exercise 6.24. Implement the steepest-descent strategy to nd the minimum of

J(x) in (6.8), modeling the program after polyconverge.m. Run the program for
dierent values of mu, N, and x(1), and answer the same questions as in Exercise
6.23.
One way to understand the behavior of steepest-descent algorithms is to plot
the error surface, which is basically a plot of the objective as a function of the
variable that is being optimized. Figure 6.17(a) displays clearly the single global
minimum of the objective function (6.4) while Figure 6.17(b) shows the many
minima of the objective function dened by (6.8). As will be clear to anyone
who has attempted Exercise 6.24, initializing within any one of the valleys causes
the algorithm to descend to the bottom of that valley. Although true steepestdescent algorithms can never climb over a peak to enter another valley (even if
the minimum there is lower), it can sometimes happen in practice when there is
a signicant amount of noise in the measurement of the downhill direction.
Essentially, the algorithm gradually descends the error surface by moving
in the (locally) downhill direction, and dierent initial estimates may lead to

119

Sampling with Automatic Gain Control

40
0.5

20

J(x)

J(x)

30

10

0.5

0
4

2
x

20 15 10 5

0
x

10

15

20

Figure 6.17 Error surfaces corresponding to (a) the objective function (6.4) and (b)
the objective function (6.8).

dierent minima. This underscores one of the limitations of steepest-descent

methodsif there are many minima, then it is important to initialize near an
acceptable one. In some problems such prior information may easily be obtained,
while in others it may be truly unknown.
The examples of this section are somewhat simple because they involve static
functions. Most applications in communication systems deal with signals that
evolve over time, and the next section applies the steepest-descent idea in a
dynamic setting to the problem of automatic gain control (AGC). The AGC
provides a simple setting in which all three of the major issues in optimization
must be addressed: setting the goal, choosing a method of solution, and verifying
that the method is successful.
Exercise 6.25. Consider performing an iterative maximization of
J(x) = 8 6|x| + 6 cos(6x)
via (6.5) with the sign on the update reversed (so that the algorithm will maximize rather than minimize). Suppose the initialization is x[0] = 0.7.
a. Assuming the use of a suitably small stepsize , determine the convergent
value of x.
b. Is the convergent value of x in part (a) the global maximum of J(x)? Justify
your answer by sketching the error surface.
Exercise 6.26. Suppose that a unimodal single-variable performance function
has only one point with zero derivative and that all points have a positive second
derivative. TRUE or FALSE: A gradient-descent method will converge to the
global minimum from any initialization.
Exercise 6.27. Consider the modulated signal
r(t) = w(t) cos(2fc t + ),

120

Chapter 6. Sampling with Automatic Gain Control

where the absolute bandwidth of the baseband message waveform w(t) is less
than fc /2. The signals x and y are generated via
x(t) = LPF{r(t) cos(2fc t + )},
y(t) = LPF{r(t) sin(2fc t + )},
where the LPF cuto frequency is fc /2.
a. Determine x(t) in terms of w(t), fc , , and .
b. Show that


d 1 2
x (t) = x(t)y(t)
d 2
using the fact that derivatives and lters commute as in (G.5).
c. Determine the values of maximizing x2 (t).
Exercise 6.28. Consider the function
J(x) = (1 |x 2|)2 .
a. Sketch J(x) for 5 x 5.
b. Analytically determine all local minima and maxima of J(x) for 5 x 5.
Hint: d|fdb(b)| = sign(f (b)) dfdb(b) where sign(a) is dened in (6.9).
c. Is J(x) unimodal as a function of x? Explain your answer.
d. Develop an iterative gradient-descent algorithm for updating x to minimize
J.
e. For an initial estimate of x = 1.2, what is the convergent value of x determined
by an iterative gradient-descent algorithm with a satisfactorily small stepsize?
f. Compute the direction (either increasing x or decreasing x) of the update
from (d) for x = 1.2.
g. Does the direction determined in part (f) point from x = 1.2 toward the convergent value of part (e)? Should it (for a correct answer to (e))? Explain your
answer.

6.7

Automatic Gain Control

Any receiver is designed to handle signals of a certain average magnitude most
eectively. The goal of an AGC is to amplify weak signals and to attenuate strong
signals so that they remain (as much as possible) within the normal operating
range of the receiver. Typically, the rate at which the gain varies is slow compared
with the data rate, though it may be fast by human standards.
The power in a received signal depends on many things: the strength of the
broadcast, the distance from the transmitter to the receiver, the direction in
which the antenna is pointed, and whether there are any geographical features
such as mountains (or tall buildings) that block, reect, or absorb the signal.

quantization levels

Sampling with Automatic Gain Control

121

Figure 6.18 The goal of the AGC is to

maintain the dynamic range of the

signal by attenuating it when it is too
large (as in (a)) and by increasing it
when it is too small (as in (b)).

(a)

Sampler

Input
r(t)
a

(b)

Output
s(kT) = s[k]

Quality
Assessment

Figure 6.19 An automatic gain control adjusts the

gain parameter a so that the average energy at

the output remains (roughly) xed, despite
uctuations in the average received energy.

While more power is generally better from the point of view of trying to decipher the transmitted message, there are always limits to the power-handling
capabilities of the receiver. Hence, if the received signal is too large (on average),
it must be attenuated. Similarly, if the received signal is weak (on average), then
it must be amplied.
Figure 6.18 shows the two extremes that the AGC is designed to avoid. In part
(a), the signal is much larger than the levels of the sampling device (indicated
by the horizontal lines). The gain must be made smaller. In part (b), the signal
is much too small to be captured eectively, and the gain must be increased.
There are two basic approaches to an AGC. The traditional approach uses analog circuitry to adjust the gain before the sampling. The more modern approach
uses the output of the sampler to adjust the gain. The advantage of the analog
method is that the two blocks (the gain and the sampling) are separate and
do not interact. The advantage of the digital adjustment is that less additional
hardware is required since the DSP is already present for other tasks.
A simple digital system for AGC adjustment is shown in Figure 6.19. The
input r(t) is multiplied by the gain a to give the normalized signal s(t). This is
then sampled to give the output s[k]. The assessment block measures s[k] and
determines whether a must be increased or decreased.
The goal is to choose a so that the power (or average energy) of s(t) is approximately equal to some specied S 2 . Since

a2 avg{r2 (t)}t=kT avg{s2 (kT )} avg{s2 [k]},

122

Chapter 6. Sampling with Automatic Gain Control

it would be ideal to choose

a2

S2
,
avg{r2 (kT )}

(6.11)

because this would imply that avg{s2 (kT )} S 2 . The averaging operation (in
this case a moving average over a block of data of size N ) is dened by
avg{x[k]} =

1
N

k


x[i]

i=kN +1

and is discussed in Appendix G in amazing detail. Unfortunately, neither the

analog input r(t) nor its power is directly available to the assessment block in
the DSP portion of the receiver, so it is not possible to directly implement (6.11).
Is there an adaptive element that can accomplish this task? As suggested at
the beginning of Section 6.5, there are three steps to the creation of a viable
optimization approach: setting a goal, choosing a solution method, and testing.
As in any real-life engineering task, a proper mathematical statement of the
goal can be tricky, and this section proposes two (slightly dierent) possibilities
for the AGC. By comparing the resulting algorithms (essentially, alternative
forms for the AGC design), it may be possible to trade o among various design
considerations.
One sensible goal is to try to minimize a simple function of the dierence
between the power of the sampled signal s[k] and the desired power S 2 . For
instance, the averaged squared error in the powers of s and S,


1 2
1
(6.12)
JLS (a) = avg
(s [k] S 2 )2 = avg{(a2 r2 (kT ) S 2 )2 },
4
4
penalizes values of a that cause s2 [k] to deviate from S 2 . This formally mimics
the parabolic form of the objective (6.4) in the polynomial minimization example
of the previous section. Applying the steepest-descent strategy yields

dJLS (a) 
,
(6.13)
a[k + 1] = a[k]
da a=a[k]
which is the same as (6.5), except that the name of the parameter has changed
from x to a. To nd the exact form of (6.13) requires the derivative of JLS (a) with
respect to the unknown parameter a. This can be approximated by swapping the
derivative and the averaging operations, as formalized in (G.12), to give
1 d avg{(a2 r2 (kT ) S 2 )2 }
dJLS (a)
=
da
4
da


1
d(a2 r2 (kT ) S 2 )2
avg
= avg{(a2 r2 (kT ) S 2 )ar2 (kT )}.
4
da
The term a2 r2 (kT ) inside the parentheses is equal to s2 [k]. The term ar2 (kT )
outside the parentheses is not directly available to the assessment mechanism,

123

Sampling with Automatic Gain Control

though it can reasonably be approximated by s2 [k]/a. Substituting the derivative

into (6.13) and evaluating at a = a[k] gives the algorithm


2
2
2 s [k]
.
(6.14)
a[k + 1] = a[k] avg (s [k] S )
a[k]
Care must be taken when implementing (6.14) that a[k] does not approach zero.
Of course, JLS (a) of (6.12) is not the only possible goal for the AGC problem.
What is important is not the exact form of the performance function, but where
the performance function has its optimal points. Another performance function
that has a similar error surface (peek ahead to Figure 6.21) is
  2

  2 2

s [k]
a r (kT )
2
2
JN (a) = avg |a|
= avg |a|
.
(6.15)
S
S
3
3
Taking the derivative gives
d avg{|a|(a2 r2 (kT )/3 S 2 )}
dJN (a)
=
da
da


d |a|(a2 r2 (kT )/3 S 2 )
avg
= avg{sgn(a[k])(s2 [k] S 2 )},
da
where the approximation arises from swapping the order of the dierentiation
and the averaging (recall (G.12)) and the derivative of | | is the signum or sign
function, which holds as long as the argument is nonzero. Evaluating this at
a = a[k] and substituting into (6.13) gives another AGC algorithm:
a[k + 1] = a[k] avg{sgn(a[k])(s2 [k] S 2 )}.

(6.16)

Consider the logic of this algorithm. Suppose that a is positive. Since S is

xed,
avg{sgn(a[k])(s2 [k] S 2 )} = avg{(s2 [k] S 2 )} = avg{s2 [k]} S 2 .
Thus, if the average energy in s[k] exceeds S 2 , a is decreased. If the average energy
in s[k] is less than S 2 , a is increased. The update ceases when avg{s2 [k]} S 2 ,
that is, where a2 S 2 /r2 , as desired. (An analogous logic applies when a is
negative.)
The two performance functions (6.12) and (6.15) dene the updates for the
two adaptive elements in (6.14) and (6.16). JLS (a) minimizes the square of the
deviation of the power in s[k] from the desired power S 2 . This is a kind of
least-square performance function (hence the subscript LS). Such squarederror objectives are common, and will reappear in phase-tracking algorithms
in Chapter 10, in clock recovery algorithms in Chapter 12, and in equalization
algorithms in Chapter 13. On the other hand, the algorithm resulting from JN (a)
has a clear logical interpretation (the N stands for naive), and the update is
simpler, since (6.16) has fewer terms and no divisions.
To experiment concretely with these algorithms, agcgrad.m provides an implementation in Matlab. It is easy to control the rate at which a[k] changes by

124

Chapter 6. Sampling with Automatic Gain Control

choice of stepsize: a larger allows a[k] to change faster, while a smaller allows
greater smoothing. Thus, can be chosen by the system designer to trade o
the bandwidth of a[k] (the speed at which a[k] can track variations in the energy
levels of the incoming signal) versus the amount of jitter or noise. Similarly, the
length over which the averaging is done (specied by the parameter lenavg)
will also inuence the speed of adaptation; longer averages imply slower-moving,
smoother estimates while shorter averages imply faster-moving, more jittery estimates.
Listing 6.6. agcgrad.m minimize the performance function J(a) = avg{|a|((1/3)a2 r2
ds)} by choice of a
n =10000;
% number o f s t e p s i n s i m u l a t i o n
vr =1.0;
% power o f th e i n p u t
r=sqrt ( v r ) * randn ( n , 1 ) ;
% g e n e r a t e random i n p u t s
ds = 0 . 1 5 ;
% d e s i r e d power o f o u t p u t
mu= 0 . 0 0 1 ;
% algorithm s t e p s i z e
l e n a v g =10;
% l e n g t h o v e r which to a v e r a g e
a=zeros ( n , 1 ) ; a ( 1 ) = 1 ;
% i n i t i a l i z e AGC p a r a m e te r
s=zeros ( n , 1 ) ;
% i n i t i a l i z e outputs
a v e c=zeros ( 1 , l e n a v g ) ;
% v e c t o r to s t o r e a v e r a g i n g terms
f o r k =1:n1
s ( k)=a ( k ) * r ( k ) ;
% n o r m a l i z e by a ( k ) & add to a v e c
a v e c =[ sign ( a ( k ) ) * ( s ( k)2 ds ) , a v e c ( 1 : l e n a v g 1 ) ] ;
a ( k+1)=a ( k)mu*mean( a v e c ) ;
% a v e r a g e a d a p t i v e update o f a ( k )
end

Typical output of agcgrad.m is shown in Figure 6.20. The gain parameter a

adjusts automatically to make the overall power of the output s roughly equal
to the specied parameter ds. Using the default values above, where the average
power of r is approximately 1, we nd that a converges to about 0.38 since
0.382 0.15 = S 2 .
The objective JLS (a) can be implemented similarly by replacing the avec
calculation inside the for loop with
a v e c =[( s ( k)2 ds ) * ( s ( k ) 2 ) / a ( k ) , a v e c ( 1 : end 1 ) ] ;

In this case, with the default values, a converges to about 0.22, which is the
value that minimizes the least-square objective JLS (a). Thus, the answer which
minimizes JLS (a) is dierent from the answer which minimizes JN (a)! More on
this later.
As it is easy to see when playing with the parameters in agcgrad.m, the
size of the averaging parameter lenavg is relatively unimportant. Even with
lenavg=1, the algorithms converge and perform approximately the same! This
is because the algorithm updates are themselves in the form of a lowpass lter.
(See Appendix G for a discussion of the similarity between averagers and lowpass
lters.) Removing the averaging from the update gives the simpler form for JN (a)
a ( k+1)=a ( k)mu* sign ( a ( k ) ) * ( s ( k)2 ds ) ;

or, for JLS (a),

a ( k+1)=a ( k)mu* ( s ( k)2 ds ) * ( s ( k ) 2 ) / a ( k ) ;

Sampling with Automatic Gain Control

125

2
1.5

Adaptive gain parameter

1
0.5
0
Amplitude

Input r(k)

Figure 6.20 An automatic gain

control adjusts the parameter a

(in the top panel) automatically
to achieve the desired output
power.

0
5
5
Output s(k)
0
5

2000

4000
6000
Iterations

8000

10000

Try them!
Perhaps the best way to formally describe how the algorithms work is to plot
the performance functions. But it is not possible to directly plot JLS (a) or JN (a),
since they depend on the data sequence s[k]. What is possible (and often leads
to useful insights) is to plot the performance function averaged over a number
of data points (also called the error surface). As long as the stepsize is small
enough and the average is long enough, the mean behavior of the algorithm will
be dictated by the shape of the error surface in the same way that the objective
function of the exact steepest-descent algorithm (for instance, the objectives
(6.4) and (6.8)) dictates the evolution of the algorithms (6.6) and (6.10).
The following code agcerrorsurf.m shows how to calculate the error surface
for JN (a). The variable n species the number of terms to average over, and tot
sums up the behavior of the algorithm for all n updates at each possible parameter value a. The average of these (tot/n) is a close (numerical) approximation
to JN (a) of (6.15). Plotting over all a gives the error surface.
Listing 6.7. agcerrorsurf.m draw the error surface for the AGC
n =10000;
% number o f s t e p s i n s i m u l a t i o n
r=randn ( n , 1 ) ;
% g e n e r a t e random i n p u t s
ds = 0 . 1 5 ;
% d e s i r e d power o f o u t p u t
range = [ 0 . 7 : 0 . 0 2 : 0 . 7 ] ;
% s p e c i f y range o f v a l u es o f a
Jagc=zeros ( s i z e ( r a n g e ) ) ; j =0;
f o r a=r a n g e
% f o r each v a l u e a
j=j +1; t o t =0;
% t ot al cost
f o r i =1:n
% over a l l p o s s i b i l i t i e s
t o t=t o t+abs ( a ) * ( ( 1 / 3 ) * a 2 * r ( i )2 ds ) ;
end
Jagc ( j )= t o t /n ;
% take average value
end

Chapter 6. Sampling with Automatic Gain Control

0.02
cost JN(a)

0.025
cost JLS(a)

126

0.015

0
0.02

0.005
0

0.6 0.4 0.2 0 0.2 0.4 0.6

Adaptive gain a

0.04

0.6 0.4 0.2 0 0.2 0.4 0.6

Adaptive gain a

Figure 6.21 The error surface for the AGC objective functions (6.12) and (6.15) each

have two minima. As long as a can be initialized with the correct (positive) sign,
there is little danger of converging to the wrong minimum.

Similarly, the error surface for JLS (a) can be plotted using
t o t=t o t + 0 . 2 5 * ( a 2 * r ( i )2 ds ) 2 ;

% e r r o r s u r f a c e f o r JLS

The output of agcerrorsurf.m for both objective functions is shown in Figure

6.21. Observe that zero (which is a critical point of the error surface) is a local
maximum in both cases. The nal converged answers (a 0.38 for JN (a) and a
0.22 for JLS (a)) occur at minima. Were the algorithm to be initialized improperly
to a negative value, then it would converge to the negative of these values. As
with the algorithms in Figure 6.17, examination of the error surfaces shows why
the algorithms converge as they do. The parameter a descends the error surface
until it can go no further.
But why do the two algorithms converge to dierent places? The facile answer
is that they are dierent because they minimize dierent performance functions.
Indeed, the error surfaces in Figure 6.21 show minima in dierent locations.
The convergent value of a 0.38 for JN (a) is explicable because 0.382 0.15 =
S 2 . The convergent value of a = 0.22 for JLS (a) is calculated in closed form in
Exercise 6.30, and this value does a good job minimizing its cost, but it has not
solved the problem of making a2 close to S 2 . Rather, JLS (a) calculates a smaller
gain that makes avg{s2 } S 2 . The minima are dierent. The moral is this: be
wary of your performance functionsthey might do what you ask.
Exercise 6.29. Use agcgrad.m to investigate the AGC algorithm.
a. What range of stepsize mu works? Can the stepsize be too small? Can the
stepsize be too large?
b. How does the stepsize mu aect the convergence rate?
c. How does the variance of the input aect the convergent value of a?
d. What range of averages lenavg works? Can lenavg be too small? Can lenavg
be too large?
e. How does lenavg aect the convergence rate?

Sampling with Automatic Gain Control

127

Exercise 6.30. Show that the value of a that achieves the minimum of JLS (a)
can be expressed as
 
S 2 k rk2
 4 .

k rk
Is there a way to use this (closed-form) solution to replace the iteration (6.14)?
Exercise 6.31. Consider the alternative objective function J(a) =
1 2 2
2
2 a (s [k]/6 S ). Calculate the derivative and implement a variation of
the AGC algorithm that minimizes this objective. How does this version
compare with the algorithms (6.14) and (6.16)? Draw the error surface for this
algorithm. Which version is preferable?
Exercise 6.32. Try initializing the estimate a(1)=-2 in agcgrad.m. Which minimum does the algorithm nd? What happens to the data record?
Exercise 6.33. Create your own objective function J(a) for the AGC problem.
Calculate the derivative and implement a variation of the AGC algorithm that
minimizes this objective. How does this version compare with the algorithms
(6.14) and (6.16)? Draw the error surface for your algorithm. Which version do
you prefer?
Exercise 6.34. Investigate how the error surface depends on the input signal.
Replace randn with rand in agcerrorsurf.m and draw the error surfaces for
JN (a) and JLS (a).

6.8

Using an AGC to Combat Fading

One of the impairments encountered in transmission systems is the degradation
due to fading, when the strength of the received signal changes in response
to changes in the transmission path. (Recall the discussion in Section 4.1.5 on
page 64.) This section shows how an AGC can be used to counteract the fading,
assuming the rate of the fading is slow, and provided the signal does not disappear
completely.
Suppose that the input consists of a random sequence undulating slowly up
and down in magnitude, as in the top plot of Figure 6.22. The adaptive AGC
compensates for the amplitude variations, growing small when the power of the
input is large, and large when the power of the input is small. This is shown
in the middle graph. The resulting output is of roughly constant amplitude, as
shown in the bottom plot of Figure 6.22.
This gure was generated using the following code:

128

Chapter 6. Sampling with Automatic Gain Control

Input r[k]

5
0
5
1.5

Adaptive gain parameter

Figure 6.22 When the signal fades

(top), the adaptive parameter
compensates (middle), allowing
the output to maintain nearly
constant power (bottom).

1
0.5
0
5

Output s[k]

0
5
0

3
Iterations

5
x104

Listing 6.8. agcvsfading.m compensating for fading with an AGC

n =50000;
r=randn ( n , 1 ) ;
env =0.75+abs ( s i n (2 * pi * [ 1 : n ] / n ) ) ;
r=r . * env ;
ds = 0 . 5 ;
a=zeros ( 1 , n ) ; a ( 1 ) = 1 ;
s=zeros ( 1 , n ) ;
mu= 0 . 0 1 ;
f o r k =1:n1
s ( k)=a ( k ) * r ( k ) ;
a ( k+1)=a ( k)mu * ( s ( k)2 ds ) ;
end

%
%
%
%
%
%
%
%

# steps in simulation
g e n e r a t e random i n p u t s
th e f a d i n g p r o f i l e
a p p l y to raw i n p u t r [ k ]
d e s i r e d power o f o u t p u t
i n i t i a l i z e AGC p a r a m e te r
i n i t i a l i z e outputs
algorithm s t e p s i z e

% n o r m a l i z e by a to g e t s
% a d a p t i v e update o f a ( k )

The fading prole dened by the vector env is slow compared with the rate
at which the adaptive gain moves, which allows the gain to track the changes.
Also, the power of the input never dies away completely. The problems that
follow ask you to investigate what happens in more extreme situations.
Exercise 6.35. Mimic the code in agcvsfading.m to investigate what happens
when the input signal dies away. (Try removing the abs command from the
fading prole variable.) Can you explain what you see?
Exercise 6.36. Mimic the code in agcvsfading.m to investigate what happens
when the power of the input signal varies rapidly. What happens if the sign of
the gain estimate is incorrect?
Exercise 6.37. Would the answers to the previous two problems change on
using algorithm (6.14) instead of (6.16)?

Sampling with Automatic Gain Control

6.9

129

Summary
Sampling transforms a continuous-time analog signal into a discrete-time digital
signal. In the time domain, this can be viewed as a multiplication by a train of
pulses. In the frequency domain this corresponds to a replication of the spectrum.
As long as the sampling rate is fast enough that the replicated spectra do not
overlap, the sampling process is reversible; that is, the original analog signal can
be reconstructed from the samples.
An AGC can be used to make sure that the power of the analog signal remains
in the region where the sampling device operates eectively. The same AGC,
when adaptive, can also provide a protection against signal fades. The AGC can
be designed using a steepest-descent (optimization) algorithm that updates the
adaptive parameter by moving in the direction of the negative of the derivative.
This steepest-descent approach to the solution of optimization problems will be
used throughout Software Receiver Design.

For Further Reading

Details about resampling procedures are available in the published works of

r Smith, J. O. Bandlimited interpolationinterpretation and algorithm,

1993, http://ccrma-www.stanford.edu/jos/resample/ .
A general introduction to adaptive algorithms centered around the steepestdescent approach can be found in

r B. Widrow and S. D. Stearns, Adaptive Signal Processing, Prentice-Hall, 1985.

One of our favorite discussions of adaptive methods is

r C. R. Johnson Jr., Lectures on Adaptive Parameter Estimation, Prentice-Hall,

1988.

Digital Filtering and the DFT

Once the received signal is sampled, the real story of the digital receiver begins.
An analog bandpass lter at the front end of the receiver removes extraneous
signals (for instance, it removes television-frequency signals from a radio receiver)
but some portion of the signal from other FDM users may remain. While it would
be conceptually possible to remove all but the desired user at the start, accurate
retunable analog lters are complicated and expensive to implement. Digital
lters, on the other hand, are easy to design, inexpensive (once the appropriate
DSP hardware is present), and easy to retune. The job of cleaning up out-ofband interferences left over by the analog BPF can be left to the digital portion
of the receiver.
Of course, there are many other uses for digital lters in the receiver, and
this chapter focuses on how to build digital lters. The discussion begins by
considering the digital impulse response and the related notion of discrete-time
convolution. Conceptually, this closely parallels the discussion of linear systems in
Chapter 4. The meaning of the DFT (discrete Fourier transform) closely parallels
the meaning of the Fourier transform, and several examples encourage uency in
the spectral analysis of discrete data signals. The nal section on practical ltering shows how to design digital lters with (more or less) any desired frequency
response by using special Matlab commands.

7.1

Discrete Time and Discrete Frequency

The study of discrete-time (digital) signals and systems parallels that of
continuous-time (analog) signals and systems. Many digital processes are fundamentally simpler than their analog counterparts, though there are a few subtleties unique to discrete-time implementations. This section begins with a brief
overview and comparison, and then proceeds to discuss the DFT, which is the
discrete counterpart of the Fourier transform.

Digital Filtering and the DFT

131

Just as the impulse function (t) plays a key role in dening signals and systems
in continuous time, the discrete pulse

1 k=0
[k] =
(7.1)
0 k = 0
can be used to decompose discrete signals and to characterize discrete-time
systems.1 Any discrete-time signal can be written as a linear combination
of discrete impulses. For instance, if the signal w[k] is the repeating pattern
{1, 1, 2, 1, 1, 1, 2, 1, . . .}, it can be written
w[k] = [k] + [k 1] + 2[k 2] + [k 3] [k 4]
+ [k 5] + 2[k 6] + [k 7] . . .
In general, the discrete-time signal w[k] can be written
w[k] =

w[j][k j].

j=

This is the discrete analog of the sifting property (4.4); simply replace the integral
with a sum, and replace (t) with [k].
Like their continuous-time counterparts, discrete-time systems map input signals into output signals. Discrete-time LTI (linear time-invariant) systems are
characterized by an impulse response h[k], which is the output of the system
when the input is an impulse, though, of course, (7.1) is used instead of (4.2).
When an input x[k] is more complicated than a single pulse, the output y[k] can
be calculated by summing all the responses to all the individual terms, and this
leads directly to the denition of discrete-time convolution:
y[k] =

x[j]h[k j] x[k] h[k].

(7.2)

j=

Observe that the convolution of discrete-time sequences appears in the reconstruction formula (6.3), and that (7.2) parallels continuous-time convolution in
(4.8) with the integral replaced by a sum and the impulse response h(t) replaced
by h[k].
The discrete-time counterpart of the Fourier transform is the discrete Fourier
transform (DFT). Like the Fourier transform, the DFT decomposes signals into
their constituent sinusoidal components. Like the Fourier transform, the DFT
provides an elegant way to understand the behavior of LTI systems by looking
at the frequency response (which is equal to the DFT of the impulse response).
Like the Fourier transform, the DFT is an invertible, information-preserving
transformation.
1

The pulse in discrete time is considerably more straightforward than the implicit denition
of the continuous-time impulse function in (4.2) and (4.3).

132

Chapter 7. Digital Filtering and the DFT

The DFT diers from the Fourier transform in three useful ways. First, it
applies to discrete-time sequences, which can be stored and manipulated directly
in computers (rather than to analog waveforms, which cannot be directly stored
in digital computers). Second, it is a sum rather than an integral, and so is easy
to implement either in hardware or in software. Third, it operates on a nite
data record, rather than an integration over all time. Given a data record (or
vector) w[k] of length N , the DFT is dened by
W [n] =

N
1


w[k]ej(2/N )nk

(7.3)

k=0

for n = 0, 1, 2, . . . , N 1. For each value n, (7.3) multiplies each term of the

data by a complex exponential, and then sums. Compare this with the Fourier
transform; for each frequency f , (2.1) multiplies each point of the waveform by
a complex exponential, and then integrates. Thus W [n] is a kind of frequency
function in the same way that W (f ) is a function of frequency. The next section
will make this relationship explicit by showing how ej(2/N )nk can be viewed as
a discrete-time sinusoid with frequency proportional to n. Just as a plot of the
frequency function W (f ) is called the spectrum of the signal w(t), plots of the
frequency function W [n] are called the (discrete) spectrum of the signal w[k].
One source of confusion is that the frequency f in the Fourier transform can
take on any value while the frequencies present in (7.3) are all integer multiples
n of a single fundamental with frequency 2/N . This fundamental is precisely
the sine wave with period equal to the length N of the window over which the
DFT is taken. Thus, the frequencies in (7.3) are constrained to a discrete set;
these are the discrete frequencies of the section title.
The most common implementation of the DFT is called the fast Fourier transform (FFT), which is an elegant way to rearrange the calculations in (7.3) so that
it is computationally ecient. For all purposes other than numerical eciency,
the DFT and the FFT are synonymous.
Like the Fourier transform, the DFT is invertible. Its inverse, the IDFT, is
dened by
w[k] =

N 1
1 
W [n]ej(2/N )nk
N n=0

(7.4)

for k = 0, 1, 2, . . . , N 1. The IDFT takes each point of the frequency function

W [n], multiplies by a complex exponential, and sums. Compare this with the
IFT; (D.2) takes each point of the frequency function W (f ), multiplies by a
complex exponential, and integrates. Thus, the Fourier transform and the DFT
translate from the time domain into the frequency domain, while the inverse
Fourier transform and the IDFT translate from frequency back into time.
Many other aspects of continuous-time signals and systems have analogs in
discrete time. Here are some that will be useful in later chapters.

Digital Filtering and the DFT

133

r Symmetryif the time signal w[k] is real, then W [n] = W [N n]. This is
analogous to (A.35).

r Parsevals theorem holds in discrete time w2 [k] = 1/N  |W [n]|2 . This

k
n
is analogous to (A.43).

r The frequency response H[n] of an LTI system is the DFT of the impulse
r
r
r
r
r
r

response h[k]. This is analogous to the continuous-time result that the frequency response H(f ) is the Fourier transform of the impulse response h(t).
Time-delay property in discrete timew[k l] W [n]ej(2/N )l . This is
analogous to (A.37).
Modulation propertythis frequency-shifting property is analogous to (A.34).
If w[k] = sin(2f k/T ) is a periodic sine wave, then the spectrum is a sum of
two delta impulses. This is analogous to the result in Example 4.1.
Convolution2 in (discrete) time is the same as multiplication in (discrete)
frequency. This is analogous to (4.10).
Multiplication in (discrete) time is the same as convolution in (discrete) frequency. This is analogous to (4.11).
The transfer function of a LTI system is the ratio of the DFT of the output
and the DFT of the input. This is analogous to (4.12).

Exercise 7.1. Show why Parsevals theorem is true in discrete time.

Exercise 7.2. Suppose a lter has impulse response h[k]. When the input is
x[k], the output is y[k]. Show that, if the input is xd [k] = x[k] x[k 1], then
the output is yd [k] = y[k] y[k 1]. Compare this result with Exercise 4.13.
Exercise 7.3. Let w[k] = sin(2k/N ) for k = 1, 2, . . . , N 1. Use the denitions
(7.3) and (7.4) to nd the corresponding values of W [n].

7.1.1

Understanding the DFT

Dene a vector W containing all N frequency values and a vector w containing
all N time values,
w = (w[0], w[1], w[2], . . . , w[N 1])T ,
W = (W [0], W [1], W [2], . . . , W [N 1])T ,
2

To be precise, this should be circular convolution. However, for the purposes of designing
a workable receiver, this distinction is not essential. The interested reader can explore the
relationship of discrete-time convolution in the time and frequency domains in a concrete
way using waystofilt.m on page 143.

134

Chapter 7. Digital Filtering and the DFT

and let M 1 be a matrix with columns of complex exponentials

1
1
1
1

1
1 ej2/N
ej4/N
ej6/N
ej2(N 1)/N

j4/N
j8/N
j12/N
e
e
ej4(N 1)/N
1 e

1 ej6/N
ej12/N
ej18/N ej6(N 1)/N .
.

..
..
..
..
.

.
.
.
.
2
j2(N 1)/N j4(N 1)/N j6(N 1)/N
j2(N 1) /N
e
e
e
1e
Then the IDFT equation (7.4) can be rewritten as a matrix multiplication
w=

1 1
M W
N

(7.5)

and the DFT is

W = N M w.

(7.6)

Since the inverse of an orthonormal matrix is equal to its own complex conjugate
transpose, M in (7.6) is the same as M 1 in (7.5) with the signs on all the
exponents ipped.
The matrix M 1 is highly structured. Let Cn be the nth column of M 1 . On
multiplying both sides by N , (7.5) can be rewritten as
N w = W [0] C0 + W [1] C1 + + W [N 1] CN 1 =

N
1


W [n]Cn .

(7.7)

n=0

This form displays the time vector w as a linear combination3 of the columns Cn .
What are these columns? They are vectors of discrete (complex-valued) sinusoids,
each at a dierent frequency. Accordingly, the DFT reexpresses the time vector
as a linear combination of these sinusoids. The complex scaling factors W [n]
dene how much of each sinusoid is present in the original signal w[k].
To see how this works, consider the rst few columns. C0 is a vector of all ones;
it is the zero-frequency sinusoid, or DC. C1 is more interesting. The ith element
of C1 is ej2i/N , which means that, as i goes from 0 to N 1, the exponential
assumes N uniformly spaced points around the unit circle. This is clearer in
polar coordinates, where the magnitude is always unity and the angle is 2i/N
radians. Thus, C1 is the lowest-frequency sinusoid that can be represented (other
than DC); it is the sinusoid that ts exactly one period in the time interval N Ts ,
where Ts is the distance in time between adjacent samples. C2 is similar, except
that the ith element is ej4i/N . Again, the magnitude is unity and the phase is
4i/N radians. Thus, as i goes from 0 to N 1, the elements are N uniformly
spaced points which go around the circle twice. Thus, C2 has frequency twice
that of C1 , and it represents a complex sinusoid that ts exactly two periods into
3

Those familiar with advanced linear algebra will recognize that M 1 can be thought of as a
change of basis that reexpresses w in a basis dened by the columns of M 1 .

Digital Filtering and the DFT

135

the time interval N Ts . Similarly, Cn represents a complex sinusoid of frequency

n times that of C1 ; it orbits the circle n times and is the sinusoid that ts exactly
n periods in the time interval N Ts .
One subtlety that can cause confusion is that the sinusoids in Ci are complexvalued, yet most signals of interest are real. Recall from Eulers identities (2.4)
and (A.3) that the real-valued sine and cosine can each be written as a sum of
two complex-valued exponentials that have exponents with opposite signs. The
DFT handles this elegantly. Consider CN 1 . This is
!
"T
2
1, ej2(N 1)/N , ej4(N 1)/N , ej6(N 1)/N , . . . , ej2(N 1) /N ,
which can be rewritten as
"T
!
1, ej2/N , ej4/N , ej6/N , . . . , ej2(N 1)/N ,
since ej2 = 1. Thus, the elements of CN 1 are identical to the elements of C1 ,
except that the exponents have the opposite sign, implying that the angle of
the ith entry in CN 1 is 2i/N radians. Thus, as i goes from 0 to N 1, the
exponential assumes N uniformly spaced points around the unit circle, in the
opposite direction from C1 . This is the meaning of what might be interpreted
as negative frequencies that show up when taking the DFT. The complex
exponential proceeds in a (negative) clockwise manner around the unit circle,
rather than in a (positive) counterclockwise direction. But it takes both to make
a real-valued sine or cosine, as Eulers formula shows. For real-valued sinusoids
of frequency 2n/N , both W [n] and W [N n] are nonzero, and they are equal
in magnitude.4
Exercise 7.4. Which column Ci represents the highest possible frequency in the
DFT? What do the elements of this column look like? Hint: look at CN/2 and
think of a square wave. This square wave is the highest frequency that can be
represented by the DFT, and occurs at exactly the Nyquist rate.

7.1.2

Using the DFT

Fortunately, Matlab makes it easy to do spectral analysis with the DFT by providing a number of simple commands that carry out the required calculations
and manipulations. It is not necessary to program the sum (7.3) or the matrix
multiplication (7.5). The single-line commands W = t(w) and w = it(W)
invoke ecient FFT (and IFFT) routines when possible, and relatively inecient DFT (and IDFT) calculations otherwise. The numerical idiosyncrasies are
completely transparent, with one annoying exception. In Matlab, all vectors,
including W and w, must be indexed from 1 to N instead of from 0 to N 1.
4

The discrete version of the symmetry property (A.35) shows that W [n] = W [N n]. Hence
the magnitudes are equal but the phases have opposite signs.

Chapter 7. Digital Filtering and the DFT

Magnitude

What do these mean?

Figure 7.1 Naive and deceptive

200

400

600
N = 210

800

1000

1200

Magnitude

136

500

1000
1500
N = 211

2000

plots of the spectrum of a sine

wave in which the frequency of
the analyzed wave appears to
depend on the size N of the
analysis window. The top
gure has N = 210 , while the
bottom uses N = 211 .

2500

While the FFT/IFFT commands are easy to invoke, their meaning is not
always instantly transparent. The intent of this section is to provide some examples that show how to interpret (and how not to interpret) the frequency-analysis
commands in Matlab.
Begin with a simple sine wave of frequency f sampled every Ts seconds, as
is familiar from previous programs such as speccos.m. The rst step in any
frequency analysis is to dene the window over which the analysis will take
place, since the FFT/DFT must operate on a nite data record. The program
specsin0.m denes the length of the analysis with the variable N (powers of
two make for fast calculations), and then analyzes the rst N samples of w. It is
tempting to simply invoke the Matlab commands fft and to plot the results.
Typing plot(fft(w(1:N))) gives a meaningless answer (try it!) because the
output of the fft command is a vector of complex numbers. When Matlab
plots complex numbers, it plots the imaginary vs. the real parts. In order to view
the magnitude spectrum, rst use the abs command, as shown in specsin0.m.
Listing 7.1. specsin0.m naive and deceptive spectrum of a sine wave via the FFT
f =100; Ts =1/1000; time = 5 . 0 ;
t=Ts : Ts : time ;
w=s i n (2 * pi * f * t ) ;
N=210;
fw=abs ( f f t (w ( 1 :N ) ) ) ;
plot ( fw )

%
%
%
%
%
%

f r e q , s a m p l i n g i n t e r v a l , time
d e f i n e a time v e c t o r
d e f i n e th e s i n u s o i d
s i z e o f a n a l y s i s window
f i n d magnitude o f DFT/FFT
p l o t th e waveform

Running this program results in a plot of the magnitude of the output of the
FFT analysis of the waveform w. The top plot in Figure 7.1 shows two large
spikes, one near 100 and one near 900. What do these mean? Try a simple
experiment. Change the value of N from 210 to 211 . This is shown in the bottom
plot of Figure 7.1, where the two spikes now occur at about 200 and at about
1850. But the frequency of the sine wave hasnt changed! It does not seem
reasonable that the window over which the analysis is done should change the
frequencies in the signal.

Magnitude

Digital Filtering and the DFT

137

Figure 7.2 Proper use of the

100

200

300

400

500

Magnitude

500 400 300 200 100 0 100 200 300 400 500
Frequency

FFT command can be done as

in specsin1.m (the top graph),
which plots only the positive
frequencies, or as in
specsin2.m (the bottom
graph), which shows the full
magnitude spectrum
symmetric about f = 0.

There are two problems. First, specsin0.m plots the magnitude data against
the index of the vector fw, and this index (by itself) is meaningless. The discussion surrounding (7.7) shows that each element of W [n] represents a scaling
of the complex sinusoid with frequency ej2n/N . Hence, these indices must be
scaled by the time over which the analysis is conducted, which involves both
the sampling interval and the number of points in the FFT analysis. The second
problem is the ordering of the frequencies. Like the columns Cn of the DFT
matrix M (7.6), the frequencies represented by the W [N n] are the negative
of the frequencies represented by W [n].
There are two solutions. The rst is appropriate only when the original signal is real-valued. In this case, the W [n]s are symmetric and there is no extra
information contained in the negative frequencies. This suggests plotting only
the positive frequencies, a strategy that is followed in specsin1.m.
Listing 7.2. specsin1.m spectrum of a sine wave via the FFT/DFT
f =100; Ts =1/1000; time = 5 . 0 ;
t=Ts : Ts : time ;
w=s i n (2 * pi * f * t ) ;
N=210;
s s f =(0:N/2 1)/( Ts *N ) ;
fw=abs ( f f t (w ( 1 :N ) ) ) ;
plot ( s s f , fw ( 1 :N/ 2 ) )

%
%
%
%
%
%
%

f r e q , s a m p l i n g i n t e r v a l , time
d e f i n e a time v e c t o r
d e f i n e th e s i n u s o i d
s i z e o f a n a l y s i s window
frequency vector
f i n d magnitude o f DFT/FFT
pl ot f o r p o s i t i v e f r eq only

The output of specsin1.m is shown in the top plot of Figure 7.2. The magnitude
spectrum shows a single spike at 100 Hz, as is expected. Change f to other values,
and observe that the location of the peak in frequency moves accordingly. Change
the width and location of the analysis window N and verify that the location of
the peak does not change. Change the sampling interval Ts and verify that the
analyzed peak remains at the same frequency.
The second solution requires more bookkeeping of indices, but gives plots
that more closely accord with continuous-time intuition and graphs. specsin2.m
exploits the built-in function fftshift, which shues the output of the FFT command so that the negative frequencies occur on the left, the positive frequencies
on the right, and DC in the middle.

138

Chapter 7. Digital Filtering and the DFT

Listing 7.3. specsin2.m spectrum of a sine wave via the FFT/DFT

f =100; Ts =1/1000; time = 1 0 . 0 ;
t=Ts : Ts : time ;
w=s i n (2 * pi * f * t ) ;
N=210;
s s f =(N/ 2 :N/2 1)/( Ts *N ) ;
fw=f f t (w ( 1 :N ) ) ;
f w s=f f t s h i f t ( fw ) ;
plot ( s s f , abs ( f w s ) )

%
%
%
%
%
%
%
%

f r e q , s a m p l i n g i n t e r v a l , time
d e f i n e a time v e c t o r
d e f i n e th e s i n u s o i d
s i z e o f a n a l y s i s window
frequency vector
do DFT/FFT
shift it for plotting
p l o t magnitude spectrum

Running this program results in the bottom plot of Figure 7.2, which shows the
complete magnitude spectrum both for positive and for negative frequencies. It
is also easy to plot the phase spectrum by substituting phase for abs in either
of the preceding two programs.
Exercise 7.5. Explore the limits of the FFT/DFT technique by choosing
extreme values. What happens in the following cases?
a. f becomes too large. Try f = 200, 300, 450, 550, 600, 800, 2200 Hz. Comment
on the relationship between f and Ts.
b. Ts becomes too large. Try Ts = 1/500, 1/250, 1/50. Comment on the relationship between f and Ts. (You may have to increase time in order to have
enough samples to operate on.)
c. N becomes too large or too small. What happens to the location in the peak
of the magnitude spectrum when N = 211 , 214 , 28 , 24 , 22 , 220 ? What happens
to the width of the peak in each of these cases? (You may have to increase
time in order to have enough samples to operate on.)
Exercise 7.6. Replace the sin function with sin2 . Use w=sin(2*pi*f*t).2. What
is the spectrum of sin2 ? What is the spectrum of sin3 ? Consider sink . What is the
largest k for which the results make sense? Explain what limitations there are.
Exercise 7.7. Replace the sin function with sinc. What is the spectrum of the
sinc function? What is the spectrum of sinc2 ?
Exercise 7.8. Plot the spectrum of w(t) = sin(t) + jet . Should you use the
technique of specsin1.m or of specsin2.m? Hint: think symmetry.
Exercise 7.9. The FFT of a real sequence is typically complex, and sometimes
it is important to look at the phase (as well as the magnitude).
a. Let w=sin(2*pi*f*t+phi). For phi = 0, 0.2, 0.4, 0.8, 1.5, 3.14, nd the phase of
the FFT output at the frequencies f.
b. Find the phase of the output of the FFT when w=sin(2*pi*f*t+phi).2.
These are all examples of simple functions, which can be investigated (in
principle, anyway) analytically. The greatest strength of the FFT/DFT is that
it can also be used for the analysis of data when no functional form is known.

Digital Filtering and the DFT

139

There is a data le on the website called gong.wav, which is a sound recording

of an Indonesian gong (a large struck metal plate). The following code reads in
the waveform and analyzes its spectrum using the FFT. Make sure that the le
gong.wav is in an active Matlab path, or you will get a le not found error.
If there is a sound card (and speakers) attached, the sound command plays the
.wav le at the sampling rate fs = 1/Ts .
Listing 7.4. specgong.m nd spectrum of the gong sound
f i l e n a m e= gong . wav ;
% name o f wave f i l e
[ x , s r ]=wavread( f i l e n a m e ) ;
% r ea d i n w a v e f i l e
Ts=1/ s r ;
% sample i n t e r v a l & # o f s a m p l e s
N=215; x=x ( 1 :N ) ;
% length f or anal ys i s
sound ( x , 1 / Ts )
% p l a y sound ( i f p o s s i b l e )
time=Ts * ( 0 : length ( x ) 1 ) ;
% time b a s e f o r p l o t t i n g
subplot ( 2 , 1 , 1 ) , plot ( time , x )
% and p l o t top f i g u r e
magx=abs ( f f t ( x ) ) ;
% t a k e FFT magnitude
s s f =(0:N/2 1)/( Ts *N ) ;
% f r e q base f o r p l o t t i n g
subplot ( 2 , 1 , 2 ) , plot ( s s f , magx ( 1 :N/ 2 ) ) % p l o t mag spectrum

Running specgong.m results in the plot shown in Figure 7.3. The top gure
shows the time behavior of the sound as it rises very quickly (when the gong
is struck) and then slowly decays over about 1.5 s. The variable N denes the
window over which the frequency analysis occurs. The lower-left plot shows the
complete spectrum, and the bottom-right plot zooms in on the low-frequency
portion where the largest spikes occur. This sound consists primarily of three
major frequencies, at about 520, 630, and 660 Hz. Physically, these represent the
three largest resonant modes of the vibrating plate.
With N = 216 , specgong.m analyzes approximately 1.5 s (Ts*N seconds, to
be precise). It is reasonable to suppose that the gong sound might undergo
important transients during the rst few milliseconds. This can be investigated
by decreasing N and applying the DFT to dierent segments of the data record.
Exercise 7.10. Determine the spectrum of the gong sound during the rst 0.1 s.
What value of N is needed? Compare this with the spectrum of a 0.1 s segment
chosen from the middle of the sound. How do they dier?
Exercise 7.11. A common practice when taking FFTs is to plot the magnitude
on a log scale. This can be done in Matlab by replacing the plot command
with semilogy. Try it in specgong.m. What extra details can you see?
Exercise 7.12. The waveform of the sound produced by another, much larger
gong is given in gong2.wav on the website. Conduct a thorough analysis of this
sound, looking at the spectrum for a variety of analysis windows (values of N)
and at a variety of times within the waveform.
Exercise 7.13. Choose a .wav le from the website (in the Sounds folder) or
download a .wav le of a song from the Internet. Conduct a FFT analysis of the
rst few seconds of sound, and then another analysis in the middle of the song.

140

Chapter 7. Digital Filtering and the DFT

Amplitude

1
0

Magnitude

0.5
1
Time in seconds

1.5

0.5

1
1.5
2
Frequency in Hertz

2.5
x104

0
400 450 500 550 600 650 700 750 800
Frequency in Hertz

Figure 7.3 Time and frequency plots of the gong waveform. The top gure shows the
decay of the signal over 1.5 s. The bottom left gure shows the magnitude spectrum,
and the right gure zooms in on the low-frequency portion so that the frequencies are
more legible.

How do the two compare? Can you correlate the FFT analysis with the pitch of
the material? With the rhythm? With the sound quality?
The key factors in a DFT- or FFT-based frequency analysis are as follows.

r The sampling interval Ts is the time resolution, the shortest time over which
r
r

any event can be observed. The sampling rate fs = 1/Ts is inversely proportional.
The total time is T = N Ts , where N is the number of samples in the analysis.
The frequency resolution is 1/T = 1/(N Ts ) = fs /N . Sinusoids closer together
(in frequency) than this value are indistinguishable.

For instance, in the analysis of the gong conducted in specgong.m, the sampling
interval Ts = 1/44100 is dened by the recording. With N = 216 , the total time
is N Ts = 1.48 s, and the frequency resolution is 1/(N Ts ) = 0.67 Hz.
Sometimes the total absolute time T is xed. Sampling faster decreases Ts and
increases N , but cannot give better resolution in frequency. Sometimes it is possible to increase the total time. Assuming a xed Ts , this implies an increase in
N and better frequency resolution. Assuming a xed N , this implies an increase
in Ts and worse resolution in time. Thus, better resolution in frequency means
worse resolution in time. Conversely, better resolution in time means worse resolution in frequency. If this is still confusing, or if you would like to see it from
a dierent perspective, check out Appendix D.
The DFT is a key tool in analyzing and understanding the behavior of communications systems. Whenever data ow through a system, it is a good idea
to plot the data as a function of time, and also as a function of frequency; that
is, to look at the data in the time domain and in the frequency domain. Often,

Digital Filtering and the DFT

141

aspects of the data that are clearer in time are hard to see in frequency, and
aspects that are obvious in frequency are obscure in time. Using both points of
view is common sense.

7.2

Practical Filtering
Filtering can be viewed as the process of emphasizing or attenuating certain
frequencies within a signal. Linear time-invariant lters are common because they
are easy to understand and straightforward to implement. Whether in discrete
or continuous time, an LTI lter is characterized by its impulse response (i.e.,
its output when the input is an impulse). The process of convolution aggregates
the impulse responses from all the input instants into a formula for the output.
It is hard to visualize the action of convolution directly in the time domain,
making analysis in the frequency domain an important conceptual tool. The
Fourier transform (or the DFT in discrete time) of the impulse response gives
the frequency response, which is easily interpreted as a plot that shows how much
gain or attenuation (or phase shift) each frequency experiences as a result of the
ltering operation. Thus, while implementing the lter in the time domain as a
convolution, it is normal to specify, design, and understand it in the frequency
domain as a point-by-point multiplication of the spectrum of the input and the
frequency response of the lter.
In principle, this provides a method not only of understanding the action
of a lter, but also of designing a lter. Suppose that a particular frequency
response is desired, say one that removes certain frequencies, while leaving others
unchanged. For example, if the noise is known to lie in one frequency band while
the important signal lies in another frequency band, then it is natural to design
a lter that removes the noisy frequencies and passes the signal frequencies.
This intuitive notion translates directly into a mathematical specication for the
frequency response. The impulse response can then be calculated directly by
taking the inverse transform, and this impulse response denes the desired lter.
While this is the basic principle of lter design, a number of subtleties can arise,
and sophisticated routines are available in Matlab that make the lter design
process exible, even if they are not foolproof.
Filters are classied in several ways.

r Lowpass lters (LPF) try to pass all frequencies below some cuto frequency
and remove all frequencies above.

r Highpass lters try to pass all frequencies above some specied value and
remove all frequencies below.

r Notch (or bandstop) lters try to remove particular frequencies (usually in a

narrow band) and to pass all others.

r Bandpass lters try to pass all frequencies in a particular range and to reject
all others.

142

Chapter 7. Digital Filtering and the DFT

The region of frequencies allowed to pass through a lter is called the passband,
while the region of frequencies removed is called the stopband. Sometimes there
is a region between where it is relatively less important what happens, and this
is called the transition band.
By linearity, more complex lter specications can be implemented as sums
and concatenations of the above basic lter types. For instance, if h1 [k] is the
impulse response of a bandpass lter that passes only frequencies between 100
and 200 Hz, and h2 [k] is the impulse response of a bandpass lter that passes
only frequencies between 500 and 600 Hz, then h[k] = h1 [k] + h2 [k] passes only
frequencies between 100 and 200 Hz or between 500 and 600 Hz. Similarly, if
hl [k] is the impulse response of a lowpass lter that passes all frequencies below
600 Hz, and hh [k] is the impulse response of a highpass lter that passes all
frequencies above 500 Hz, then h[k] = hl [k] hh [k] is a bandpass lter that passes
only frequencies between 500 and 600 Hz, where represents convolution.
Filters are also classied by the length of their impulse response. If the output
of a lter depends only on a specied number of samples, the lter is said to
have a nite impulse response, abbreviated FIR. Otherwise, it is said to have an
innite impulse response, abbreviated IIR. The bulk of the lters in Software
Receiver Design are FIR lters with a at passband, because these are the
most common lters in a typical receiver. But other lter proles are possible,
and the techniques of lter design are not restricted to at passbands. Section
7.2.1 shows several ways that digital FIR lters can be implemented in Matlab.
IIR lters arise whenever there is a loop (such as in a phase-locked loop), and
one special case (the integrator) is an intrinsic piece of the adaptive elements.
Section 7.2.2 shows how to implement IIR lters. Section 7.2.3 shows how to
design lters with specic properties, and how they behave on a number of test
signals.

7.2.1

Implementing FIR Filters

Suppose that the impulse response of a discrete-time lter is h[i], i =
0, 1, 2, . . . , N 1. If the input to the lter is the sequence x[i], i = 0, 1, . . . , M 1,
then the output is given by the convolution Equation (7.2). There are four ways
to implement this ltering in Matlab.

r conv directly implements the convolution equation and outputs a vector of

length N + M 1.

r filter implements the convolution so as to supply one output value for each
input value; the output is of length M .

r In the frequency domain, take the FFT of the input and the FFT of the
output, multiply the two, and take the IFFT to return to the time domain.

r In the time domain, pass through the input data, at each time multiplying by
the impulse response and summing the result.
Probably the easiest way to see the dierences is to play with the four methods.

Digital Filtering and the DFT

143

Listing 7.5. waystofilt.m conv vs. lter vs. freq domain vs. time domain
h=[1 1 2 2 3 3 ] ;
% impulse response h [ k ]
x =[1 2 3 4 5 6 5 4 3 2 1 ] ;
% i n p u t data x [ k ]
yconv=conv ( h , x )
% convolve x [ k ] * h [ k ]
y f i l t =f i l t e r ( h , 1 , x )
% f i l t e r x [ k ] with h [ k ]
n=length ( h)+length ( x ) 1;
% pad l e n g t h f o r FFT
f f t h=f f t ( [ h zeros ( 1 , nlength ( h ) ) ] ) ;
% FFT o f h [ k ] i s H[ n ]
f f t x=f f t ( [ x , zeros ( 1 , nlength ( x ) ) ] ) ;
% FFT o f i n p u t i s X[ n ]
f f t y=f f t h . * f f t x ;
% p r o d u c t o f H[ n ] and X[ n ]
y f r e q=r e a l ( i f f t ( f f t y ) )
% IFFT o f p r o d u c t g i v e s y [ k ]
z =[ zeros ( 1 , length ( h ) 1) , x ] ;
% i n i t i a l f i l t e r s t a t e =0
f o r k =1: length ( x )
% time domain method
ytim ( k)= f l i p l r ( h ) * z ( k : k+length ( h ) 1 ) ; % do f o r each x [ k ]
end
% to d i r e c t l y c a l c u l a t e y [ k ]

Observe that the rst M terms of yconv, yfilt, yfreq, and ytim are the same,
but that both yconv and yfreq have N-1 extra values at the end. For both the
time-domain method and the filter command, the output values are aligned
in time with the input values, one output for each input. Eectively, the filter
command is a single-line implementation of the time-domain for loop.
For the FFT method, the two vectors (input and convolution) must both have
length N+M-1. The raw output has complex values due to numerical roundo,
and the command real is used to strip away the imaginary parts. Thus, the
FFT-based method requires more Matlab commands to implement. Observe
also that conv(h,x) and conv(x,h) are the same, whereas filter(h,1,x) is
not the same as filter(x,1,h).
To view the frequency response of the lter h, Matlab provides the command
freqz, which automatically zero pads5 the impulse response and then plots both
the magnitude and the phase. Type
freqz (h)

to see that the lter with impulse response h=[1, 1, 1, 1, 1] is a (poor)

lowpass lter with two dips at 0.4 and 0.8 of the Nyquist frequency as shown in
Figure 7.4. The command freqz always normalizes the frequency axis so that
1.0 corresponds to the Nyquist frequency fs /2. The passband of this lter
(all frequencies less than the point where the magnitude drops 3 dB below the
maximum) ends just below 0.2. The maximum magnitude in the stopband occurs
at about 0.6, where it is about 12 dB down from the peak at zero. Better (i.e.,
closer to the ideal) lowpass lters would attenuate more in the stopband, would
be atter across the passband, and would have narrower transition bands.
5

By default, the Matlab command freqz creates a vector of length 512 that contains the
specied impulse response followed by zeros. The FFT of this elongated vector is used for
the magnitude and phase plots, giving the plots a smoother appearance than when taking
the FFT of the raw impulse response.

Chapter 7. Digital Filtering and the DFT

Magnitude (dB)

144

20
0
20

Figure 7.4 The frequency

response of the lter with

impulse response h=[1, 1,
1, 1, 1] has a poor lowpass
character. It is easier to see
this in the frequency domain
than directly in the time
domain.

40

Phase (degrees)

100
0

100
0

7.2.2

0.2
0.4
0.6
0.8
Normalized frequency (Nyquist = 1)

Implementing IIR Filters

At rst glance it might seem counterintuitive that a useful lter could have an
impulse response that is innitely long. To see how this inniteness might arise
in a specic case, suppose that the output y[k] of an LTI system is created from
the input x[k] according to the formula
y[k + 1] = ay[k] + bx[k],

(7.8)

where a and b are constants. This is shown in the left-hand portion of Figure
7.5, where the time delay between successive terms (between k and k + 1 for
instance) is represented by z 1 . This is an example of the Z-transform, and is
described more fully in Appendix F. If the initial value is y[0] = 0, the response
of the system (7.8) to an impulse x[k] = [k] (where [k] is the discrete impulse
function (7.1)) is
b, ba, ba2, ba3 , ba4 , . . . .

(7.9)

If |a| > 1, this impulse response increases towards innity and the system is
said to be unstable. If |a| < 1, the values in the sequence die away towards (but
never quite reaching) zero, and the system is stable. In both cases, the impulse
response is innitely long and so the system (7.8) is IIR. The special case when
a = 1 is called a summer (or, in analogy with continuous time, an integrator)

because y[k] = y[0] + b k1
i=0 x[i] sums up all the inputs. The summer is often
represented symbolically as in the right-hand part of Figure 7.5.
The general form of an IIR lter is
y[k + 1] = a0 y[k] + a1 y[k 1] + + an y[k n]
+ b0 x[k] + b1 x[k 1] + + bm x[k m],
which can be rewritten more concisely as
y[k + 1] = aT y[k] + bT x[k]

Digital Filtering and the DFT

x[k]
b

Delay

y[k+1]

z1

x[k]

y[k]

145

y[k]

a
Figure 7.5 The LTI system y[k + 1] = ay[k] + bx[k], with input x[k] and output y[k], is

represented schematically using the delay z 1 in this simple block diagram. The
special case when a = 1 is called a summer because it eectively adds up all the input
values, and it is often drawn more concisely as on the right side of the gure.

where
a = [a0 , a1 , . . . , an ]T
b = [b0 , b1 , . . . , bm ]T
are column vectors of the parameters of the lter and where
x[k] = [x[k], x[k 1], . . . , x[k m]]T
y[k] = [y[k], y[k 1], . . . , y[k n]]T
are vectors of past input and output data. The symbol z T indicates the transpose
of the vector z. In Matlab, transpose is represented using the single apostrophe
, but take care: when using complex numbers, the apostrophe actually implements a complex conjugate as well as transpose.
Matlab has a number of functions that make it easy to explore IIR ltering.
For example, the function impz plots the impulse response of a discrete IIR (or
FIR) lter. The impulse response of the system (7.8) is found by entering
b=1;
a =[1 , 0. 8] ;
impz ( b , a )

The parameters are the vectors b=[b0 , b1 , . . . , bm ] and a=[1, a0 , a1 , . . . , an ].

Hence the above code gives the response when a of (7.8) is +0.8. Similarly, the
command freqz(b,a) displays the frequency response for the IIR lter, where
b and a are constructed in the same way.
The routine waystofiltIIR.m explores ways of implementing IIR lters. The
simplest method uses the filter command. For instance filter(b,a,d) lters
the data d through the IIR system with parameters b and a dened as above.
This gives the same output as rst calculating the impulse response h and then
using filter(h,1,d). The fastest (and most general) way to implement the
lter is in the for loop at the end, which mimics the time-domain method for
FIR lters, but includes the additional parameters a.
Listing 7.6. waystofiltIIR.m ways to implement IIR lters
a =[1 0 . 8 ] ; l e n a=length ( a ) 1;
b = [ 1 ] ; l e n b=length ( b ) ;
d=randn ( 1 , 2 0 ) ;

% autoregressive co efficients
% moving a v e r a g e c o e f f i c i e n t s
% data to f i l t e r

146

Chapter 7. Digital Filtering and the DFT

i f l e n a >=l en b ,
h=impz ( b , a ) ;
y f i l t =f i l t e r ( h , 1 , d )
end
y f i l t 2 =f i l t e r ( b , a , d )
y=zeros ( l e n a , 1 ) ; x=zeros ( l en b , 1 ) ;
f o r k =1: length ( d) l e n b
x=[d ( k ) ; x ( 1 : l e n b 1 ) ] ;
ytim ( k)=a ( 2 : l e n a +1) * y+b * x ;
y=[ ytim ( k ) ; y ( 1 : l e n a 1 ) ] ;
end

% d i m p u l s e n e e d s l e n a >=l e n b
% impulse response of f i l t e r
% f i l t e r x [ k ] with h [ k ]
%
%
%
%
%
%

f i l t e r u s i n g a and b
i n i t i a l states in f i l t e r
time domain method
past values of inputs
directly calculate y[ k ]
past values of outputs

Like FIR lters, IIR lters can be lowpass, bandpass, and highpass, or can
have almost any imaginable eect on the spectrum of their input.
Exercise 7.14. Some IIR lters can be used as inverses to some FIR lters.
Show that the FIR lter
1
a
x[k] = r[k] r[k 1]
b
b
is an inverse of the IIR lter (7.8).
Exercise 7.15. FIR lters can be used to approximate the behavior of IIR lters
by truncating the impulse response. Create a FIR lter with impulse response
given by the rst 10 terms of (7.9) for a = 0.9 and b = 2. Simulate the FIR lter
and the IIR lter (7.8) in Matlab, using the same random input to both. Verify
that the outputs are (approximately) the same.

7.2.3

Filter Design
This section gives an extended explanation of how to use Matlab to design a
bandpass lter to t a specied frequency response with a at passband. The
same procedure (with suitable modication) also works for the design of other
basic lter types.
A bandpass lter is intended to scale, but not distort, signals with frequencies that fall within the passband, and to reject signals with frequencies in the
stopband. An ideal, distortionless response for the passband would be perfectly
at in magnitude, and would have linear phase (corresponding to a delay). The
transition band from the passband to the stopband should be as narrow as possible. In the stopband, the frequency response magnitude should be suciently
small and the phase is of no concern. These objectives are captured in Figure
7.6. Recall (from (A.35)) for a real w(t) that |W (f )| is even and  W (f ) is odd,
as illustrated in Figure 7.6.
Matlab has several commands that carry out lter design. The firpm command6 provides a linear phase impulse response (with real, symmetric coecients
6

Some early versions of Matlab use remez instead of firpm.

Digital Filtering and the DFT

Transition
region

147

Transition
region

|H( f )|
Passband
Stopband

Figure 7.6 Specications of a

Lower
H( f ) band
edge

Upper
band
edge

real-valued bandpass lter in terms

of magnitude and phase spectra.

h[k]) that has the best approximation of a specied (piecewise-at) frequency

response.7 The syntax of the firpm command for the design of a bandpass lter
as in Figure 7.6 is
b = f i r p m ( f l , f b e , damps ) ,

which has inputs fl, fbe, and damps, and output b.

r fl species (one fewer than) the number of terms in the impulse response

r
r
7

of the desired lter. Generally, more is better in terms of meeting the design
specications. However, larger fl are also more costly in terms of computation
and in terms of the total throughput delay, so a compromise is usually made.
fbe is a vector of frequency band edge values as a fraction of the prevailing
Nyquist frequency. For example, the lter specied in Figure 7.6 needs six
values: the bottom of the stopband (presumably zero), the top edge of the
lower stopband (which is also the lower edge of the lower transition band),
the lower edge of the passband, the upper edge of the passband, the lower edge
of the upper stopband, and the upper edge of the upper stopband (generally
the last value will be 1). The transition bands must have some nonzero width
(the upper edge of the lower stopband cannot equal the lower passband edge)
or Matlab produces an error message.
damps is the vector of desired amplitudes of the frequency response at each
band edge. The length of damps must match the length of fbe.
b is the output vector containing the impulse response of the specied lter.
There are many possible meanings of the word best; for the firpm algorithm, best is
dened in terms of maintaining an equal ripple in the at portions.

Magnitude Response (dB)

Chapter 7. Digital Filtering and the DFT

Phase (degrees)

148

20
0
20

Figure 7.7 Frequency response

40

of the bandpass lter designed

by the firpm command in
bandex.m. The variable fbe
species a set of frequencies (as
a fraction of the Nyquist rate)
and damps species the
corresponding amplitudes. The
freqz command plots both the
magnitude and phase spectra
of the lter.

60
80
500
0
500
1000

1500
0

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

Normalized frequency (Nyquist = 1)

The following Matlab script designs a lter that mimics Figure 7.6:
Listing 7.7. bandex.m design a bandpass lter and plot frequency response
f b e =[0 0 . 2 4 0 . 2 6 0 . 7 4 0 . 7 6 1 ] ;
damps=[0 0 1 1 0 0 ] ;
f l =30;
b=f i r p m ( f l , f b e , damps ) ;
figure (1)
freqz (b)

%
%
%
%
%

f r e q band e d g e s a s a f r a c t i o n o f
th e N y q u i s t f r e q u e n c y
d e s i r e d a m p l i t u d e s a t band e d g e s
f i lt e r size
b i s th e d e s i g n e d i m p u l s e r e s p o n s e

% p l o t f r e q r e s p o n s e to c h e c k d e s i g n

The frequency response of the resulting nite impulse response (FIR) lter is
shown in Figure 7.7. Observe that the stopband is about 14 dB lower than the
passband, a marginal improvement over the naive lowpass lter of Figure 7.4,
but the design is much atter in the passband. The equiripple nature of this
lter is apparent in the slow undulations of the magnitude in the passband.
Designing IIR lters is also straightforward. For example, the code
[ b , a ]= cheby1 ( 1 0 , 5 , 0 . 2 5 ) ;

designs a 10th-order lowpass lter with 5 dB of ripple in the passband, which

extends from DC to 0.25 times the Nyquist rate.
While commands such as firpm and cheby1 make lter design easy, be
warnedstrange things can happen, even to nice lters. Whenever using any
of the lter design tools, always check freqz(b,a) to verify that the output of
the design behaves as anticipated. There are many other ways to design LTI
lters, and Matlab includes several commands that design lter coecients:
cfirpm, firls, fir1, fir2, butter, cheby2, and ellip. The subject of lter
design is vast, and each of these is useful in certain applications. For simplicity, we have chosen to present most examples throughout Software Receiver
Design by using firpm.

Digital Filtering and the DFT

149

Exercise 7.16. Rerun bandex.m with very narrow transition regions, for
instance fbe = [0, 0.24, 0.2401, 0.6, 0.601, 1]. What happens to the ripple in
the passband? Compare the minimum magnitude in the passband with the maximum value in the stopband.
Exercise 7.17. Returning to the lter specied in Figure 7.6, try using dierent
numbers of terms in the impulse response, fl = 5, 10, 100, 500, 1000. Comment
on the resulting designs in terms of atness of the frequency response in the
passband, attenuation from the passband to the stopband, and the width of the
transition band.
Exercise 7.18. Specify and design a lowpass lter with cuto at 0.15. What
values of fl, fbe, and damps work best?
Exercise 7.19. Specify and design a lter that has two passbands, one between
[0.2, 0.3] and another between [0.5, 0.6]. What values of fl, fbe, and damps work
best?
Exercise 7.20. Use the lter designed in bandex.m to lter a whitenoise sequence (generated by randn) using the time-domain method from
waystofilt.m. Verify that the output has a bandpass spectrum.
The preceding lter designs do not explicitly require the sampling rate of the
signal. However, since the sampling rate determines the Nyquist rate, it is used
implicitly. The next exercise asks that you familiarize yourself with real units
of frequency in the lter design task.
Exercise 7.21. In Exercise 7.10, the program specgong.m was used to analyze the sound of an Indonesian gong. The three most prominent partials (or
narrowband components) were found to be at about 520, 630, and 660 Hz.
a. Design a lter using firpm that will remove the two highest partials from this
sound without aecting the lowest partial.
b. Use the filter command to process the gong.wav le with your lter.
c. Take the FFT of the resulting signal (the output of your lter) and verify that
the partial at 520 remains while the others are removed.
d. If a sound card is attached to your computer, compare the sound of the raw
and the ltered gong sound by using Matlabs sound command. Comment
on what you hear.
The next set of problems examines how accurate digital lters really are.
Exercise 7.22. With a sampling rate of 44100 Hz, let x[k] be a sinusoid of
frequency 3000 Hz. Design a lowpass lter with a cuto frequency fl of 1500
Hz, and let y[k] = LPF{x[k]} be the output of the lter.

150

Chapter 7. Digital Filtering and the DFT

a. How much does the lter attenuate the signal? (Express your answer as the
ratio of the power in the output y[k] to the power in the input x[k].)
b. Now use an LPF with a cuto of 2500 Hz. How much does the lter attenuate
the signal?
c. Now use an LPF with a cuto of 2900 Hz. How much does the lter attenuate
the signal?
Exercise 7.23. Repeat Exercise 7.22 without using the filter command
(implement the ltering, using the time-domain method in waystofilt.m).
Exercise 7.24. With the same setup as in Exercise 7.22, generate x[k] as a
bandlimited noise signal containing frequencies between 3000 Hz and the Nyquist
rate.
a. Use an LPF with cuto frequency fl of 1500 Hz. How much does the lter
attenuate the signal?
b. Now use an LPF with a cuto of 2500 Hz. How much does the lter attenuate
the signal?
c. Now use an LPF with a cuto of 3100 Hz. How much does the lter attenuate
the signal?
d. Now use an LPF with a cuto of 4000 Hz. How much does the lter attenuate
the signal?
Exercise 7.25. Let f1 < f2 < f3 . Suppose x[k] has no frequencies above f1 Hz,
while z[k] has no frequencies below f3 . If an LPF has cuto frequency f2 , in
principle,
LPF{x[k] + z[k]} = LPF{x[k]} + LPF{z[k]} = x[k] + 0 = x[k].
Explain how this is (and is not) consistent with the results of Exercises 7.22
and 7.24.
Exercise 7.26. Let the output y[k] of an LTI system be created from the input
x[k] according to the formula
y[k + 1] = y[k] + x[k],
where is a small constant. This is drawn in Figure 7.5.
a. What is the impulse response of this lter?
b. What is the frequency response of this lter?
c. Would you call this lter lowpass, highpass, or bandpass?
Exercise 7.27. Using one of the alternative lter design routines (cfirpm,
firls, fir1, fir2, butter, cheby1, cheby2, or ellip), repeat Exercises 7.16
7.21. Comment on the subtle (and the not-so-subtle) dierences in the resulting
designs.

Digital Filtering and the DFT

151

Exercise 7.28. The eect of bandpass ltering can be accomplished by

a. modulating to DC,
b. lowpass ltering, and
c. modulating back.
Repeat the task given in Exercise 7.21 (the Indonesian gong lter design problem)
by modulating with a 520 Hz cosine, lowpass ltering, and then remodulating.
Compare the nal output of this method with the direct bandpass lter design.

For Further Reading

Here are some of our favorite books on signal processing:

r K. Steiglitz, A Digital Signal Processing Primer, Addison-Wesley, 1996;

r J. H. McClellan, R. W. Schafer, and M. A. Yoder, DSP First: A Multimedia
Approach, Prentice-Hall, 1998;

r C. S. Burrus and T. W. Parks, DFT/FFT and Convolution Algorithms:

Theory and Implementation, Wiley-Interscience, 1985.

Bits to Symbols to Signals

Any message, whether analog or digital, can be translated into a string of binary
digits. In order to transmit or store these digits, they are often clustered or
encoded into a more convenient representation whose elements are the symbols
of an alphabet. In order to utilize bandwidth eciently, these symbols are then
translated (again!) into short analog waveforms called pulse shapes that are combined to form the actual transmitted signal.
The receiver must undo each of these translations. First, it examines the
received analog waveform and decodes the symbols. Then it translates the symbols back into binary digits, from which the original message can, with luck, be
reconstructed.
This chapter briey examines each of these translations, and the tools needed
to make the receiver work. One of the key ideas is correlation, which can be used
as a kind of pattern-matching tool for discovering key locations within the signal
stream. Section 8.3 shows how correlation can be viewed as a kind of linear lter,
and hence its properties can be readily understood both in the time and in the
frequency domain.

8.1

Bits to Symbols
The information that is to be transmitted by a communication system comes in
many forms: a pressure wave in the air, a ow of electrons in a wire, a digitized
image or sound le, the text in a book. If the information is in analog form, then
it can be sampled (as in Chapter 6). For instance, an analog-to-digital converter
can transform the output of a microphone into a stream of numbers representing
the pressure wave in the air, or it can turn measurements of the current in the wire
into a sequence of numbers that are proportional to the electron ow. The sound
le, which is already digital, contains a long list of numbers that correspond to
the instantaneous amplitude of the sound. Similarly, the picture le contains a
list of numbers that describe the intensity and color of the pixels in the image.

Bits to Symbols to Signals

153

The text can be transformed into a numerical list using the ASCII code. In all
these cases, the raw data represent the information that must be transmitted by
the communication system. The receiver, in turn, must ultimately translate the
received signal back into the data.
Once the information is encoded into a sequence of numbers, it can be reexpressed as a string of binary digits 0 and 1. This is discussed at length in
Chapter 14. But the binary 01 representation is not usually very convenient
from the point of view of ecient and reliable data transmission. For example,
directly modulating a binary string with a cosine wave would result in a small
piece of the cosine wave for each 1 and nothing (the zero waveform) for each 0.
It would be very hard to tell the dierence between a message that contained a
string of zeros, and no message at all!
The simplest solution is to recode the binary 0, 1 into binary 1. This can be
accomplished using either the linear operation 2x 1 (which maps 0 into 1,
and 1 into 1), or 2x + 1 (which maps 0 into 1, and 1 into 1). This binary
1 is an example of a two-element symbol set. There are many other common
symbol sets. In multilevel signaling, the binary terms are gathered into groups.
Regrouping in pairs, for instance, recodes the information into a four-level signal.
For example, the binary sequence might be paired thus:
. . . 000010110101 . . . . . . 00 00 10 11 01 01 . . . .

(8.1)

Then the pairs might be encoded as

11
10
01
00

+3
+1
1
3

(8.2)

to produce the symbol sequence

. . . 00 00 10 11 01 01 . . . . . . 3, 3, +1, +3 , 1, 1 . . . .
Of course, there are many ways in which such a mapping between bits and
symbols might be made, and Exercise 8.2 explores one simple alternative called
the Gray code. The binary sequence may be grouped in many ways: into triplets
for an 8-level signal, into quadruplets for a 16-level scheme, into in-phase and
quadrature parts for transmission through a quadrature system. The values
assigned to the groups (1, 3 in (8.2)) are called the alphabet of the given
system.

Example 8.1. Text is commonly encoded using ASCII, and Matlab automatically represents any string le as a list of ASCII numbers. For instance, let
str=I am text be a text string. This can be viewed in its internal form by
typing real(str), which returns the vector 73 32 97 109 32 116 101 120 116,
which is the (decimal) ASCII representation of this string. This can be viewed

154

Chapter 8. Bits to Symbols to Signals

in binary using dec2base(str,2,8), which returns the binary (base 2) representation of the decimal numbers, each with eight digits.
The Matlab function letters2pam.m, provided on the website, changes a
text string into the four-level alphabet 1, 3. Each letter is represented by a
sequence of four elements, for instance the letter I is 1, 3, 1, 1. The function is invoked with the syntax letters2pam(str). The inverse operation is
pam2letters.m. Thus pam2letters(letters2pam(str)) returns the original
string.
One complication in the decoding procedure is that the receiver must gure
out when the groups begin in order to parse the digits properly. For example,
if the rst element of the sequence in (8.1) were lost, the message would be
mistranslated as
. . . 00010110101 . . . . . .00 01 01 10 10 . . .
. . . 3, 1, 1, 1, 1, . . . .
Similar parsing problems occur whenever messages start or stop. For example,
if the message consists of pixel values for a television image, it is important that
the decoder be able to determine precisely when the image scan begins. These
kinds of synchronization issues are typically handled by sending a special start
of frame sequence that is known both to the transmitter and to the receiver.
The decoder then searches for the start sequence, usually using some kind of
correlation (pattern-matching) technique. This is discussed in detail in Section
8.3.

Example 8.2. There are many ways to translate data into binary equivalents. Example 8.1 showed one way to convert text into 4-PAM and then into
binary. Another way exploits the Matlab function text2bin.m and its inverse
bin2text.m, which use the 7-bit version of the ASCII code (rather than the
8-bit version). This representation is more ecient, since each pair of text letters
can be represented by 14 bits (or seven 4-PAM symbols) rather than 16 bits (or
eight 4-PAM symbols). On the other hand, the 7-bit version can encode only
half as many characters as the 8-bit version. Again, it is important to be able to
correctly identify the start of each letter when decoding.
Exercise 8.1. The Matlab code in naivecode.m, which is on the website,
implements the translation from binary to 4-PAM (and back again) suggested in
(8.2). Examine the resiliency of this translation to noise by plotting the number
of errors as a function of the noise variance v. What is the largest variance for
which no errors occur? At what variance are the errors near 50%?
Exercise 8.2. A Gray code has the property that the binary representation for
each symbol diers from its neighbors by exactly one bit. A Gray code for the

Bits to Symbols to Signals

155

translation of binary into 4-PAM is

01
11
10
00

+3
+1
1
3

Mimic the code in naivecode.m to implement this alternative and plot the number of errors as a function of the noise variance v. Compare your answer with
Exercise 8.1. Which code is better?

8.2

Symbols to Signals
Even though the original message is translated into the desired alphabet, it is
not yet ready for transmission: it must be turned into an analog waveform. In the
binary case, a simple method is to use a rectangular pulse of duration T seconds
to represent +1, and the same rectangular pulse inverted (i.e., multiplied by 1)
to represent the element 1. This is called a polar non-return-to-zero line code.
The problem with such simple codes is that they use bandwidth ineciently.
Recall that the Fourier transform of the rectangular pulse in time is the sinc(f )
function in frequency (A.20), which dies away slowly as f increases. Thus, simple
codes like the non-return-to-zero are compact in time, but wide in frequency,
limiting the number of simultaneous nonoverlapping users in a given spectral
band.
More generally, consider the four-level signal of (8.2). This can be turned into
an analog signal for transmission by choosing a pulse shape p(t) (that is not
necessarily rectangular and not necessarily of duration T ) and then transmitting
p(t kT ) if
p(t kT ) if
3p(t kT ) if
3p(t kT ) if

the
the
the
the

kth
kth
kth
kth

symbol
symbol
symbol
symbol

is 1
is 1
is 3
is 3

Thus, the sequence is translated into an analog waveform by initiating a scaled

pulse at the symbol time kT , where the amplitude scaling is proportional to the
associated symbol value. Ideally, the pulse would be chosen so that

r the value of the message at time k does not interfere with the value of the
r
r

message at other sample times (the pulse shape causes no intersymbol interference),
the transmission makes ecient use of bandwidth, and
the system is resilient to noise.

Unfortunately, these three requirements cannot all be optimized simultaneously,

and so the design of the pulse shape must consider carefully the tradeos that are
needed. The focus in Chapter 11 is on how to design the pulse shape p(t), and the

156

Chapter 8. Bits to Symbols to Signals

1
0.8
0.6
0.4
0.2
0

0.1

10

0.2

0.3

0.4
0.5
0.6
The pulse shape

0.7

0.8

0.9

20
30
40
50
60
70
80
90
The waveform representing Transmit this text

100

3
2
1
0
1
2
3

Figure 8.1 The process of pulse shaping replaces each symbol of the alphabet (in this
case, 1, 3) with an analog pulse (in this case, the short blip function shown in the
top panel).

consequences of that choice in terms of possible interference between adjacent

symbols and in terms of the signal-to-noise properties of the transmission.
For now, to see concretely how pulse shaping works, lets pick a simple nonrectangular shape and proceed without worrying about optimality. Let p(t) be
the symmetric blip shape shown in the top part of Figure 8.1, and dened in
pulseshape.m by the hamming command. The text string in str is changed into
a four-level signal as in Example 8.1, and then the complete transmitted waveform is assembled by assigning an appropriately scaled pulse shape to each data
value. The output appears at the bottom of Figure 8.1. Looking at this closely,
observe that the rst letter T is represented by the four values 1, 1, 1, and
3, which correspond exactly to the rst four negative blips, three small and
one large.
The program pulseshape.m represents the continuous-time or analog signal
by oversampling both the data sequence and the pulse shape by a factor of
M. This technique was discussed in Section 6.3, where an analog sine wave
sine100hzsamp.m was represented digitally at two sampling intervals, a slow
digital interval Ts and a faster rate (shorter interval) Ts /M representing the
underlying analog signal. The pulse shaping itself is carried out by the filter
command which convolves the pulse shape with the data sequence.
Listing 8.1. pulseshape.m applying a pulse shape to a text string
s t r= Transmit this text string ; % message to be t r a n s m i t t e d
m=l e t t e r s 2 p a m ( s t r ) ; N=length (m) ; % 4 l e v e l s i g n a l o f l e n g t h N

Bits to Symbols to Signals

157

M=10; mup=zeros ( 1 ,N*M) ; mup ( 1 :M:N*M)=m; % o v e r s a m p l e by M

ps=hamming (M) ;
% b l i p p u l s e o f width M
x=f i l t e r ( ps , 1 , mup ) ;
% c o n v o l v e p u l s e shape & data

Exercise 8.3. Assume the sampling period is T = 0.1, and then plot the spectrum of the output x. What is the bandwidth of this signal?
Exercise 8.4. Again with T = 0.1, plot the spectrum of the output x when the
pulse shape is changed to a rectangular pulse. (Change the denition of ps in
the next-to-last line of pulseshape.m.) What is the bandwidth of this signal?
Exercise 8.5. Can you think of a pulse shape that will have a narrower bandwidth than either of the above but that will still be time limited by T ? Implement
it by changing the denition of ps, and check to see whether you are correct.
Thus the raw message is prepared for transmission by

r encoding into an alphabet (in this case 1, 3), and then

r pulse shaping the elements of the alphabet using p(t).
The receiver must undo these two operations; it must examine the received signal
and recover the elements of the alphabet, and then decode these to reconstruct
the message. Both of these tasks are made easier using correlation, which is
discussed in the next section. The actual decoding processes used in the receiver
are then discussed in Section 8.4.

8.3

Correlation
Suppose there are two signals or sequences. Are they similar, or are they dierent? If one is just shifted in time relative to the other, how can the time shift
be determined? The approach called correlation shifts one of the sequences in
time, and calculates how well they match (by multiplying point by point and
summing) at each shift. When the sum is small, they are not much alike; when
the sum is large, many terms are similar. Thus, correlation is a simple form of
pattern matching, which is useful in communication systems for aligning signals
in time. This can be applied at the level of symbols when it is necessary to nd
appropriate sampling times, and it can be applied at the frame level when it
is necessary to nd the start of a message (for instance, the beginning of each
frame of a television signal). This section discusses various techniques of crosscorrelation and autocorrelation, which can be viewed in either the time domain
or the frequency domain.

158

Chapter 8. Bits to Symbols to Signals

In discrete time, cross-correlation is a function of the time shift j between two

sequences w[k] and v[k + j]:
1
Rwv (j) = lim
T T

T /2


w[k]v[k + j].

(8.3)

k=T /2

For nite data records, the sum need only be accumulated over the nonzero
elements, and the normalization by 1/T is often ignored. (This is how
Matlabs xcorr function works.) While this may look like the convolution
Equation (7.2), it is not identical since the indices are dierent (in convolution, the index of v() is j k instead of k + j). The operation and meaning
of the two processes are also not identical: convolution represents the manner in which the impulse response of a linear system acts on its inputs to
give the outputs, while cross-correlation quanties the similarity of two signals.
In many communication systems, each message is parcelled into segments or
frames, each having a predened header. As the receiver decodes the transmitted
message, it must determine where the message segments start. The following
code simulates this in a simple setting in which the header is a predened binary
string and the data consist of a much longer binary string that contains the
header hidden somewhere inside. After performing the correlation, the index
with the largest value is taken as the most likely location of the header.
Listing 8.2. correx.m correlation can locate the header within the data
h e a d e r =[1 1 1 1 1 1 1 1 1 1 ] ;
% header i s p r ed e f i n e d s t r i n g
l o c =30; r =25;
% p l a ce header i n p o s i t i o n l o c
data =[ sign ( randn ( 1 , l o c 1)) h e a d e r sign ( randn ( 1 , r ) ) ] ; % make s i g n a l
sd = 0 . 2 5 ; data=data+sd * randn ( s i z e ( data ) ) ;
% add n o i s e
y=x c o r r ( header , data ) ;
% do c r o s s c o r r e l a t i o n
[m, i n d ]=max( y ) ;
% find largest correlation
h e a d s t a r t=length ( data ) i n d +1;
% p l a c e where h ea d er s t a r t s

Running correx.m results in a trio of gures much like those in Figure 8.2.
(Details will dier each time it is run, because the actual data are randomly
generated with Matlabs randn function.) The top plot in Figure 8.2 shows the
10-sample binary header. The data vector is constructed to contain l = 30 data
values followed by the header (with noise added), and then r = 25 more data
points, for a total block of 65 points. It is plotted in the middle of Figure 8.2.
Observe that it is dicult to see where the header lies among the noisy data
record. The correlation between the data and the header is calculated and plotted
at the bottom of Figure 8.2 as a function of the lag index. The index for which
the correlation attains its largest value denes where the best match between the
data and the header occurs. Most likely this will be at index ind=35 (as in Figure
8.2). Because of the way Matlab orders its output, the calculations represent
sliding the rst vector (the header), term by term, across the second vector (the

Bits to Symbols to Signals

159

Header
1
0.5
0
0.5
1

10

Data with embedded header

2
1
0
1
2

10

20

30

40

50

60

70

10
Correlation of header with data
5
0
5
10

20

40

60

80

100

120

140

Figure 8.2 The correlation can be used to locate a known header within a long signal.

The predened header is shown in the top graph. The data consist of a random
binary string with the header embedded and noise added. The bottom plot shows the
correlation. The location of the header is determined by the peak occurring at 35.

data). The long string of zeros at the end1 occurs because the two vectors are of
dierent lengths. The start of the header is given by length(data)-ind+1.
One way in which the correlation might fail to nd the correct location of
the header is if the header string accidently occurred in the data values. If this
happened, then the correlation would be as large at the accidental location
as at the intended location. This becomes increasingly unlikely as the header
is made longer, though a longer header also wastes bandwidth. Another way to
decrease the likelihood of false hits is to average over several headers.
Exercise 8.6. Rerun correx.m with dierent-length data vectors (try loc=100,
r=100 and loc=10, r=10). Observe how the location of the peak changes.
Exercise 8.7. Rerun correx.m with dierent-length headers. Does the peak in
the correlation become more or less distinct as the number of terms in the header
increases?
1

Some versions of Matlab use a dierent convention with the xcorr command. If you nd
that the string of zeros occurs at the beginning, try reversing the order of the arguments.

160

Chapter 8. Bits to Symbols to Signals

Exercise 8.8. Rerun correx.m with dierent amounts of noise. Try sd=0, 0.1,
0.3, 0.5, 1, 2. How large can the noise be made if the correlation is still to
nd the true location of the header?
Exercise 8.9. The code in corrvsconv.m explores the relationship between
the correlation and convolution. The convolution of two sequences is essentially
the same as the cross-correlation of the time-reversed signal, except that the
correlation is padded with extra zeros. (The Matlab function fliplr carries
out the time reversal.) If h is made longer than x, what needs to be changed so
that yconv and ycorr remain equal?
Listing 8.3. corrvsconv.m comparing correlation and convolution
h=[1 1 2 2 3 3 ] ;
x =[1 2 3 4 5 6 5 4 3 2 1 ] ;
yconv=conv ( x , h )
y c o r r=x c o r r ( h , f l i p l r ( x ) )

8.4

%
%
%
%

def i ne sequence h [ k ]
def i ne sequence x [ k ]
convolve x [ k ] * h [ k ]
c o r r e l a t e f l i p p e d x and h

Receive Filtering: From Signals to Symbols

Suppose that a message has been coded into its alphabet, pulse-shaped into
an analog signal, and transmitted. The receiver must then un-pulse-shape the
analog signal back into the alphabet, which requires nding where in the received
signal the pulse shapes are located. Correlation can be used to accomplish this
task, because it is eectively the task of locating a known sequence (in this case
the sampled pulse shape) within a longer sequence (the sampled received signal).
This is analogous to the problem of nding the header within the received signal,
although some of the details have changed. While optimizing this procedure is
somewhat involved (and is therefore postponed until Chapter 11), the gist of the
method is reasonably straightforward, and is shown by continuing the example
begun in pulseshape.m.
The code in recfilt.m below begins by repeating the pulse-shaping code
from pulseshape.m, using the pulse shape ps dened in the top plot of Figure
8.1. This creates an analog signal x that is oversampled by a factor M . The
receiver begins by correlating the pulse shape with the received signal, using the
xcorr function.2 After appropriate scaling, this is downsampled to the symbol
rate by choosing one out of each M (regularly spaced) samples. These values
are then quantized to the nearest element of the alphabet using the function
quantalph (which was introduced in Exercise 3.25). The function quantalph
has two vector arguments; the elements of the rst vector are quantized to the
2

Because of the connections between cross-correlation, convolution, and ltering, this process
is often called pulse-matched ltering because the impulse response of the lter is matched
to the shape of the pulse.

Bits to Symbols to Signals

161

nearest elements of the second vector (in this case quantizing z to the nearest
elements of [3, 1, 1, 3]).
If all has gone well, the quantized output mprime should be identical to the
original message string. The function pam2letters.m rebuilds the message from
the received signal. The nal line of the program calculates how many symbol
errors have occurred (how many of the 1, 3 dier between the message m and
the reconstructed message mprime).
Listing 8.4. recfilt.m undo pulse shaping using correlation
y=x c o r r ( x , ps ) ;
z=y (N*M:M:N*M) / ( pow ( ps ) *M) ;
mprime=q u a n t a l p h ( z , [ 3 , 1 , 1 , 3 ] ) ;
p a m 2 l e t t e r s ( mprime )
sum( abs ( sign ( mprimem) ) )

%
%
%
%
%
%

run p u l s e s h a p e .m & c r e a t e x
co r r e l at e pulse & si gnal
downsample to symbol r a t e
q u a n t i z e to a l p h a b e t
r e c o n s t r u c t message
c a l c u l a t e number o f e r r o r s

In essence, pulseshape.m from page 156 is a transmitter, and recfilt.m is the

corresponding receiver. Many of the details of this simulation can be changed and
the message will still arrive intact. The following exercises encourage exploration
of some of the options.
Exercise 8.10. Other pulse shapes may be used. Try
a. a sinusoidal shaped pulse ps=sin(0.1*pi*(0:M1));
b. a sinusoidal shaped pulse ps=cos(0.1*pi*(0:M1));
c. a rectangular pulse shape ps=ones(1,M);
Exercise 8.11. What happens if the pulse shape used at the transmitter diers
from the pulse shape used at the receiver? Try using the original pulse shape
from pulseshape.m at the transmitter, but using the following.
a.
b.

ps=sin(0.1*pi*(0:M1)); at the receiver. What percentage errors occur?

ps=cos(0.1*pi*(0:M1)); at the receiver. What percentage errors occur?

Exercise 8.12. The received signal might not always arrive at the
receiver unchanged. Simulate a noisy channel by including the command
x=x+1.0*randn(size(x)) before the xcorr command in recfilt.m. What percentage errors occur? What happens as you increase or decrease the amount of noise
(by changing the 1.0 to a larger or smaller number)?

8.5

Frame Synchronization: From Symbols to Bits

In many communication systems, the data in the transmitted signal are separated
into chunks called frames. In order to correctly decode the text at the receiver, it
is necessary to locate the boundary (the start) of each chunk. This was done by
at in the receiver of recfilt.m by correctly indexing into the received signal

162

Chapter 8. Bits to Symbols to Signals

y. Since this starting point will not generally be known beforehand, it must
somehow be located. This is an ideal job for correlation and a marker sequence.
The marker is a set of predened symbols embedded at some specied location
within each frame. The receiver can locate the marker by cross-correlating it with
the incoming signal stream. What makes a good marker sequence? This section
shows that not all markers are created equally.
Consider the binary data sequence
. . . + 1, 1, +1, +1, 1, 1, 1, +1, M, +1, 1, +1, . . .,

(8.4)

where the marker M is used to indicate a frame transition. A seven-symbol

marker is to be used. Consider two candidates:

r marker A: 1, 1, 1, 1, 1, 1, 1
r marker B: 1, 1, 1, 1, 1, 1, 1
The correlation of the signal with each of the markers can be performed as
indicated in Figure 8.3.
For marker A, correlation corresponds to a simple sum of the last seven values.
Starting at the location of the seventh value available to us in the data sequence
(two data points before the marker), marker A produces the sequence
1, 1, 1, 1, 1, 3, 5, 7, 7, 7, 5, 5.
For marker B, starting at the same point in the data sequence and performing
the associated moving weighted sum produces
1, 1, 3, 1, 5, 1, 1, 1, 7, 1, 1, 3.
With the two correlator output sequences shown, started two values prior to
the start of the seven-symbol marker, we want the ag indicating a frame start
to occur with point number 9 in the correlator sequences shown. Clearly, the
correlator output for marker B has a much sharper peak at its ninth value than
does the correlator output of marker A. This should enhance the robustness of
the use of marker B relative to that of marker A against the unavoidable presence
of noise.
Marker B is a maximum-length pseudo-noise (PN) sequence. One property
of a maximum-length PN sequence {ci } of plus and minus ones is that its autocorrelation is quite peaked:

N 1
1 
1
k = N,
cn cn+k =
Rc (k) =
1/N k = N.
N
n=0

Another technique that involves the chunking of data and the need to locate
boundaries between chunks is called scrambling. Scrambling is used to whiten
a message sequence (to make its spectrum atter) by decorrelating the message.
The transmitter and receiver agree on a binary scrambling sequence s that is
repeated over and over to form a periodic string S that is the same size as the
message. S is then added (using modulo 2 arithmetic) bit by bit to the message

Bits to Symbols to Signals

m1

m2

m3

m4

m5

m6

m7

m1

m2

m3

m4

m5

m6

m7

m1

m1

m2

m3

m4

m5

m6

m7

m1

m2

m1

m2

m3

m4

m5

m6

m7

m1

m2

m3

m4

m5

m6

m7

m1

m2

m3

m4

m5

m6

m7

m1

m2

m3

m4

m5

m6

m7

sum = m1 m2 + m3 + m4 m5 m6 m7

m1

m2

m3

m4

m5

m6

m7

163

Data sequence with

embedded marker

Sum products of
corresponding values
1

Shift marker to
right and repeat

Shift marker to
right and repeat

Shift marker to
right and repeat

sum = m1 + m2 + m3 m4 m5 m6 + m7

m2

m3

m4

m5

m6

m7

sum = m1 + m2 m3 m4 m5 + m6 + m1m7

m3

m4

m5

m6

m7

sum = m1 m2 m3 m4 + m5 + m1m6 + m2m7

Shift marker to
right and repeat

sum = m1m1 + m2m2 + m3m3 + m4m4 + m5m5 + m6m6 + m7m7

Figure 8.3 The binary signal from (8.4) is shown with a seven-symbol embedded

marker m1 , m2 , . . ., m7 . The process of correlation multiplies successive elements of

the signal by the marker, and sums. The marker shifts to the right and the process
repeats. The sum is largest when the marker is aligned with itself because all the
terms are positive, as shown in the bottom part of the diagram. Thus the shift with
the largest correlation sum points towards the embedded marker in the data
sequence. Correlation can often be useful even when the data are noisy.

m at the transmitter, and then S is added bit by bit again at the receiver. Since
both 1 + 1 = 0 and 0 + 0 = 0,
m+S+S =m
and the message is recaptured after the two summing operations. The scrambling
sequence must be aligned so that the additions at the receiver correspond to the

164

Chapter 8. Bits to Symbols to Signals

appropriate additions at the transmitter. The alignment can be accomplished

using correlation.
Exercise 8.13. Redo the example of this section, using Matlab.
Exercise 8.14. Add a channel with impulse response 1, 0, 0, a, 0, 0, 0, b to this
example. (Convolve the impulse response of the channel with the data sequence.)
a. For a = 0.1 and b = 0.4, how does the channel change the likelihood that the
correlation correctly locates the marker? Try using both markers A and B.
b. Answer the same question for a = 0.5 and b = 0.9.
c. Answer the same question for a = 1.2 and b = 0.4.
Exercise 8.15. Generate a long sequence of binary random data with the marker
embedded every 25 points. Check that marker A is less robust (on average) than
marker B by counting the number of times marker A misses the frame start
compared with the number of times marker B misses the frame start.
Exercise 8.16. Create your own marker sequence, and repeat the previous problem. Can you nd one that does better than marker B?
Exercise 8.17. Use the 4-PAM alphabet with symbols 1, 3. Create a marker
sequence, and embed it in a long sequence of random 4-PAM data. Check to
make sure it is possible to correctly locate the markers.
Exercise 8.18. Add a channel with impulse response 1, 0, 0, a, 0, 0, 0, b to this
4-PAM example.
a. For a = 0.1 and b = 0.4, how does the channel change the likelihood that the
correlation correctly locates the marker?
b. Answer the same question for a = 0.5 and b = 0.9.
Exercise 8.19. Choose a binary scrambling sequence s that is 17 bits long.
Create a message that is 170 bits long, and scramble it using bit-by-bit mod 2
addition.
a. Assuming the receiver knows where the scrambling begins, add s to the scrambled data and verify that the output is the same as the original message.
b. Embed a marker sequence in your message. Use correlation to nd the marker
and to automatically align the start of the scrambling.

Stu Happens

There is nothing new in this chapter. Really. By peeling away the outer, most
accessible layers of the communication system, the previous chapters have provided all of the pieces needed to build an idealized digital communication system,
and this chapter just shows how to combine the pieces into a functioning system.
Then we get to play with the system a bit, asking a series of what if questions.
In outline, the idealized system consists of two parts, rather than three, since
the channel is assumed to be noiseless and disturbance-free.
The Transmitter

r codes a message (in the form of a character string) into a sequence of symbols,
r transforms the symbol sequence into an analog signal using a pulse shape, and
r modulates the scaled pulses up to the passband.
The Digital Receiver

r
r
r
r
r
r

samples the received signal,

demodulates to baseband,
lters the signal to remove unwanted portions of the spectrum,
correlates with the pulse shape to help emphasize the peaks of the pulse
train,
downsamples to the symbol rate, and
decodes the symbols back into the character string.

Each of these procedures is familiar from earlier chapters, and you may have
already written Matlab code to perform them. It is time to combine the elements into a full simulation of a transmitter and receiver pair that can function
successfully in an ideal setting.

166

Chapter 9. Stu Happens

9.1

An Ideal Digital Communication System

The system is illustrated in the block diagram of Figure 9.1. This system is
described in great detail in Section 9.2, which also provides a Matlab version
of the transmitter and receiver. Once everything is pieced together, it is easy to
verify that messages can be sent reliably from transmitter to receiver.
Unfortunately, some of the assumptions made in the ideal setting are unlikely
to hold in practice; for example, the presumption that there is no interference
from other transmitters, that there is no noise, that the gain of the channel is
always unity, that the signal leaving the transmitter is exactly the same as the
signal at the input to the digital receiver. All of these assumptions will almost
certainly be violated in practice. Stu happens!
Section 9.3 begins to accommodate some of the nonidealities encountered in
real systems by addressing the possibility that the channel gain might vary with
time. For example, a large metal truck might abruptly move between a cell phone
and the antenna at the base station, causing the channel gain to drop precipitously. If the receiver cannot react to such a change, it may suer debilitating
errors when reconstructing the message. Section 9.3 examines the eectiveness of
incorporating an automatic gain control (AGC) adaptive element (as described
in Section 6.8) at the front end of the receiver. With care, the AGC can accommodate the varying gain. The success of the AGC is encouraging. Perhaps there
are simple ways to compensate for other common impairments.
Section 9.4 presents a series of what if questions concerning the various
assumptions made in the construction of the ideal system, focusing on perfor-

Message
character
string

T-spaced
symbol
sequence

Baseband
signal

Passband
signal

Pulse
filter

Coder

cos(2fct)
(a) Transmitter

Received
signal

Ts-spaced
passband
signal

Ts-spaced
baseband
signal
Lowpass
filter

kTs
k = 0, 1, 2, ...

MTs-spaced
soft
decisions

Pulse
correlator
filter

Downsampler

Mixer

MTs-spaced
hard
decisions

Quantizer
n(MTs) + lTs
n = 0, 1, 2, ...

cos(2fckTs)
Sampler

Mixer

Demodulator
(b) Receiver
Figure 9.1 Block diagram of an ideal communication system.

Recovered
character
string

Decoder

Stu Happens

167

mance degradations caused by synchronization loss and various kinds of distortions.

What if there is channel noise? (The ideal system is noise-free.)
What if the channel has multipath interference? (There are no reections or
echoes in the ideal system.)
What if the phase of the oscillator at the transmitter is unknown (or guessed
incorrectly) at the receiver? (The ideal system knows the phase exactly.)
What if the frequency of the oscillator at the transmitter is o just a bit from
its specication? (In the ideal system, the frequency is known exactly.)
What if the sample instant associated with the arrival of top-dead-center of the
leading pulse is inaccurate so that the receiver samples at the wrong times?
(The sampler in the ideal system is never fooled.)
What if the number of samples between symbols assumed by the receiver is
dierent from that used at the transmitter? (These are the same in the ideal
case.)
These questions are investigated via a series of experiments that require only
modest modication of the ideal system simulation. These simulations will show
(as with the time-varying channel gain) that small violations of the idealized
assumptions can often be tolerated. However, as the operational conditions
become more severe (as more stu happens), the receiver must be made more
robust.
Of course, it is not possible to x all these problems in one chapter. Thats
what the rest of the book is for!
Chapter 10 deals with methods to acquire and track changes in the carrier phase
and frequency.
Chapter 11 describes better pulse shapes and corresponding receive lters that
perform well in the presence of channel noise.
Chapter 12 discusses techniques for tracking the symbol clock so that the samples
can be taken at the best possible times.
Chapter 13 designs a symbol-spaced lter that undoes multipath interference
and can reject certain kinds of in-band interference.
Chapter 14 describes simple coding schemes that provide protection against
channel noise.

9.2

Simulating the Ideal System

The simulation of the digital communication system in Figure 9.1 divides into
two parts just as the gure does. The rst part creates the analog transmitted
signal, and the second part implements the discrete-time receiver.
The message consists of the character string

168

Chapter 9. Stu Happens

01234 I wish I were an Oscar Meyer wiener 56789

In order to transmit this important message, it is rst translated into the 4-PAM
symbol set 1, 3 (which is designated m[i] for i = 1, 2, . . . , N ) using the subroutine letters2pam.m. This can be represented formally as the analog pulse
N 1
m[i](t iT ), where T is the time interval between symbols. The
train i=0
simulation operates with an oversampling factor M , which is the speed at which
the analog portion of the system evolves. The pulse train enters a lter with
pulse shape p(t). The sifting property (A.56) shows that the output of the pulseN 1
m[i]p(t iT ), which is then modulated
shaping lter is the analog signal i=0
(by multiplication with a cosine at the carrier frequency fc ) to form the transmitted signal
N
1


m[i]p(t iT )cos(2fc t).

i=0

Since the channel is assumed to be ideal, this is equal to the received signal r(t).
This ideal transmitter is simulated in the rst part of idsys.m.
Listing 9.1. idsys.m (part 1) idealized transmission system - the transmitter
% e n c o d e t e x t s t r i n g a s Ts p a c e d 4PAM s e q u e n c e
s t r= 01234 I wish I were an Oscar Meyer wiener 56789 ;
m=l e t t e r s 2 p a m ( s t r ) ; N=length (m) ;
% 4 l e v e l s i g n a l o f l e n g t h N
% z e r o pad Ts p a c e d symbol s e q u e n c e to c r e a t e upsampled
% T/Ms p a c e d s e q u e n c e o f s c a l e d Ts p a c e d p u l s e s (T=1)
M=100;
% oversampling f a c t o r
mup=zeros ( 1 ,N*M) ;
% Hamming p u l s e f i l t e r with
mup ( 1 :M:N*M)=m;
% T/Ms p a ced i m p u l s e r e s p o n s e
p=hamming (M) ;
% b l i p p u l s e o f width M
x=f i l t e r ( p , 1 , mup ) ;
% c o n v o l v e p u l s e shape with data
f i g u r e ( 1 ) , p l o t s p e c ( x , 1 /M)
% baseband AM m o d u l a ti o n
t =1/M: 1 /M: length ( x )/M;
% T/Ms p a c e d time v e c t o r
f c =20;
% c a r r i e r frequency
c=cos (2 * pi * f c * t ) ;
% carrier
r=c . * x ;
% modulate message with c a r r i e r

Since Matlab cannot deal directly with analog signals, the analog signal
r(t) is sampled at M times the symbol rate, and r(t)|t=kTs (the signal r(t)
sampled at times t = kTs ) is the vector r in the Matlab code. The vector r is
also the input to the digital portion of the receiver. Thus, the rst sampling block
in the receiver of Figure 9.1 is implicit in the way Matlab emulates the analog
signal. To be specic, k can be represented as the sum of an integer multiple of
M and some positive integer smaller than M such that
kTs = (iM + )Ts .
Since T = M Ts ,
kTs = iT + Ts .

Stu Happens

169

Thus, the received signal sampled at t = kTs is

r(t)|t=kTs =

N
1


m[i]p(t iT )cos(2fc t)|t=kTs =iT +Ts

i=0

N
1


m[i]p(kTs iT )cos(2fc kTs ).

i=0

The receiver performs downconversion in the second part of idsys.m with a

mixer that uses a synchronized cosine wave, followed by a lowpass lter that
removes out-of-band signals. A quantizer makes hard decisions that are then
decoded back from symbols to the characters of the message. When all goes
well, the reconstructed message is the same as the original.
Listing 9.2. idsys.m (part 2) idealized transmission system - the receiver
% AM d em o d u l a ti o n o f r e c e i v e d s i g n a l s e q u e n c e r
c2=cos (2 * pi * f c * t ) ;
% s y n c h r o n i z e d c o s i n e f o r mixing
x2=r . * c2 ;
% demod r e c e i v e d s i g n a l
f l =50; f b e =[0 0 . 1 0 . 2 1 ] ;
% LPF p a r a m e te r s
damps=[1 1 0 0 ] ;
b=f i r p m ( f l , f b e , damps ) ;
% c r e a t e LPF i m p u l s e r e s p o n s e
x3=2* f i l t e r ( b , 1 , x2 ) ;
% LPF and s c a l e s i g n a l
% e x t r a c t upsampled p u l s e s u s i n g c o r r e l a t i o n implemented
% a s a c o n v o l v i n g f i l t e r ; f i l t e r with p u l s e and n o r m a l i z e
y=f i l t e r ( f l i p l r ( p ) / ( pow ( p ) *M) , 1 , x3 ) ;
% s e t d e l a y to f i r s t symbolsample and i n c r e m e n t by M
z=y ( 0 . 5 * f l +M:M:N*M) ;
% downsample to symbol r a t e
f i g u r e ( 2 ) , plot ( [ 1 : length ( z ) ] , z , . ) % p l o t s o f t d e c i s i o n s
% d e c i s i o n d e v i c e and symbol matching p e r f o r m a n c e a s s e s s m e n t
mprime=q u a n t a l p h ( z , [ 3 , 1 , 1 , 3 ] ) ;
% quantize alphabet
c v a r =(mprimez ) * ( mprimez ) / length ( mprime ) , % c l u s t e r v a r i a n c e
lmp=length ( mprime ) ;
p e r e r r =100 *sum( abs ( sign ( mprimem( 1 : lmp ) ) ) ) / lmp , % symbol e r r o r
% decode d e c i s i o n d ev i ce output to tex t s t r i n g
r e c o n s t r u c t e d m e s s a g e=p a m 2 l e t t e r s ( mprime )

This ideal system simulation is composed primarily of code recycled from

previous chapters. The transformation from a character string to a four-level T spaced sequence to an upsampled (T /M -spaced) T -wide (Hamming) pulse-shape
lter output sequence mimics pulseshape.m from Section 8.2. This sequence of
T /M -spaced pulse lter outputs and its magnitude spectrum are shown in Figure
9.2 (type plotspec(x,1/M) after running idsys.m).
Each pulse is one time unit long, so successive pulses can be initiated without
any overlap. The unit duration of the pulse could be a millisecond (for a pulse
frequency of 1 kHz) or a microsecond (for a pulse frequency of 1 MHz). The
magnitude spectrum in Figure 9.2 has little apparent energy outside bandwidth
2 (the meaning of 2 in Hz is dependent on the units of time).
This oversampled waveform is upconverted by multiplication with a sinusoid.
This is familiar from AM.m of Section 5.2. The transmitted passband signal and its
spectrum (created using plotspec(v,1/M)) are shown in Figure 9.3. The default
carrier frequency is fc=20. Nyquist sampling of the received signal occurs as long

Magnitude

Amplitude

Chapter 9. Stu Happens

3
2
1
0
1
2
3

Figure 9.2 The transmitter of

40

80
120
Seconds

160

200

8000
6000
4000

idsys.m creates the message

signal x in the top plot by
convolving the 4-PAM data in m
with the pulse shape p. The
magnitude spectrum X(f ) is
shown at the bottom.

2000

Amplitude

0
50 40 30 20 10 0 10
Frequency

Magnitude

170

3
2
1
0
1
2
3

20

40

60

20

30

40

50

Figure 9.3 The received signal r

and its spectrum R(f ) after
upconversion.

80 100 120 140 160 180

Seconds

4000
3000
2000
1000
0
50 40 30 20 10 0 10
Frequency

20

30

40

50

as the sample frequency 1/(T /M ) = M for T = 1 is twice the highest frequency

in the received signal, which will be the carrier frequency plus the baseband
signal bandwidth of approximately 2. Thus, M should be greater than 44 to
prevent aliasing of the received signal. This allows reconstruction of the analog
received signal at any desired point, which could prove valuable if the times at
which the samples were taken were not synchronized with the received pulses.
The transmitted signal reaches the receiver portion of the ideal system in
Figure 9.1. Downconversion is accomplished by multiplying the samples of the
received signal by an oscillator that (miraculously) has the same frequency and
phase as was used in the transmitter. This produces a signal with the spectrum
shown in Figure 9.4 (type plotspec(x2,1/M) after running idsys.m), a spectrum that has substantial nonzero components (that must be removed) at about
twice the carrier frequency.

Magnitude response (dB)

Magnitude

Amplitude

Stu Happens

3
2
1
0
1
2
3
0

Figure 9.4 Downconversion in

20

40

60

idsys.m mixes the received

signal r with a cosine wave c2 to
generate x2. The gure shows x2
in time in the top plot and its
spectrum in the bottom plot.

80 100 120 140 160 180

Seconds

4000
3000
2000
1000
0
50 40 30 20 10 0 10
Frequency

20

30

40

50

20
0
20
40
60
80
100

Figure 9.5 The frequency

response of the lowpass lter

designed using firpm in idsys.m.

Phase (degrees)

200
400
600
800
1000

171

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

Normalized frequency (Nyquist = 1)

To suppress the components centered around 40 in Figure 9.4 and to pass the
baseband component without alteration (except for possibly a delay), the lowpass
lter is designed with a cuto near 10. For M = 100, the Nyquist frequency is 50.
(Section 7.2.3 details the use of firpm for lter design.) The frequency response
of the resulting FIR lter (from freqz(b)) is shown in Figure 9.5. To make
sense of the horizontal axes, observe that the 1 in Figure 9.5 corresponds to
the 50 in Figure 9.4. Thus the cuto between normalized frequencies 0.1 and
0.2 corresponds to an unnormalized frequency near 10, as desired.
The output of the lowpass lter in the demodulator is a signal with the spectrum shown in Figure 9.6 (drawn using plotspec(x3,1/M)). The spectrum in
Figure 9.6 should compare quite closely to that of the transmitter baseband in
Figure 9.2, as indeed it does. It is easy to check the eectiveness of the lowpass lter design by attempting to use a variety of dierent lowpass lters, as
suggested in Exercise 9.4.
Recall the discussion in Section 8.4 of matching two scaled pulse shapes. Viewing the pulse shape as a kind of marker, it is reasonable to correlate the pulse

Magnitude

Amplitude

Chapter 9. Stu Happens

3
2
1
0
1
2
3

Figure 9.6 After demodulation

20

40

60

and lowpass ltering, the signal

x3 in idsys.m is approximately
the same as the baseband
transmitted signal (and
spectrum) in Figure 9.2.

80 100 120 140 160 180

Seconds

8000
6000
4000
2000
0
50 40 30 20 10 0 10
Frequency

20

30

40

50

Best times to take samples

Amplitude of received signal

172

Delay

Figure 9.7 The rst four symbol

2
1
0
1
2
3
0
0

50

100 150 200 250 300 350 400

Ts-spaced samples
T

2T
3T
T-spaced samples

periods (recall that the oversampling

factor was M = 100) of the signal at
the receiver (after the demodulation,
LPF, and pulse correlator lter). The
rst three symbol values are 3, +3,
3, which can be deciphered from
the signal, assuming that the delay
can be selected appropriately.

4T

shape with the received signal in order to locate the pulses. (More justication for this procedure is forthcoming in Chapter 11.) This appears in Figure
9.1 as the block labelled pulse correlation lter. The code in idsys.m implements this using the filter command to carry out the correlation (rather than
the xcorr function), though the choice was a matter of convenience. (Refer to
corrvsconv.m in Exercise 8.9 to see how the two functions are related.)
The rst 4M samples of the resulting signal y are plotted in Figure 9.7 (via
plot(y(1:4*M))). The rst three symbols of the message (i.e., m(1:3)) are 3,
3, and 3, and Figure 9.7 shows why it is best to take the samples at indices
125 + kM . The initial delay of 125 corresponds to half the length of the lowpass
lter (0.5*fl) plus half the length of the correlator lter (0.5*M) plus half a
symbol period (0.5*M), which accounts for the delay from the start of each pulse
to its peak.

Stu Happens

173

3
2
1
0

Figure 9.8 Repeatedly overlaying a

time width of four symbols yields

an eye diagram.

2
3
0

50

100

150

200

250

300

350

400

Selecting this delay and the associated downsampling are accomplished in the
code
z=y ( 0 . 5 * f l +M:M: end ) ;

% downsample to symbol r a t e

in idsys.m, which recovers the T -spaced samples z. With reference to Figure

9.1, the parameter l in the downsampling block is 125.
A revealing extension of Figure 9.7 is to plot the oversampled waveform y for
the complete transmission in order to see whether the subsequent peaks of the
pulses occur at regular intervals precisely on source alphabet symbol values, as
we would hope. However, even for small messages (such as the wiener jingle),
squeezing such a gure onto one graph makes a detailed visual examination of the
peaks fruitless. This is precisely why we plotted Figure 9.7to see the detailed
timing information for the rst few symbols.
One idea is to plot the next four symbol periods on top of the rst four by
shifting the start of the second block to time zero. Continuing this approach
throughout the data record mimics the behavior of a well-adjusted oscilloscope
that triggers at the same point in each symbol group. This operation can be
implemented in Matlab by rst determining the maximum number of groups
of 4*M samples that t inside the vector y from the lth sample on. Let
u l=f l o o r ( ( length ( y) l 1)/(4 *M) ) ;

then the reshape command can be used to form a matrix with 4*M rows and ul
columns. This is easily plotted using
plot ( reshape ( y ( l : u l * 4 *M+124) ,4 *M, u l ) )

and the result is shown in Figure 9.8. Note that the rst element plotted in Figure
9.8 is the lth element of Figure 9.7. This type of gure, called an eye diagram, is
commonly used in practice as an aid in troubleshooting. Eye diagrams will also
be used routinely in succeeding chapters.
Four is an interesting grouping size for this particular problem because four
symbols are used to represent each character in the coding and decoding implemented in letters2pam.m and pam2letters.m. One idiosyncrasy is that each
character starts o with a negative symbol. Another is that the second symbol in

174

Chapter 9. Stu Happens

4
3
2
Figure 9.9 Values of the

reconstructed symbols, called the

soft decisions, are plotted in this
constellation diagram time
history.

0
1
2
3
4

20

40

60

80 100 120 140 160 180 200

each character is never 1 in our chosen message. These are not generic eects;
they are a consequence of the particular coding and message used in idsys.m.
Had we chosen to implement a scrambling scheme (recall Exercise 8.19) the
received signal would be whitened and these particular peculiarities would not
occur.
The vector z contains estimates of the decoded symbols, and the command
plot([1:length(z)],z,.) produces a time history of the output of the downsampler, as shown in Figure 9.9. This is called the time history of a constellation
diagram in which all the dots are meant to lie near the allowable symbol values.
Indeed, the points in Figure 9.9 cluster tightly about the alphabet values 1 and
3. How tightly they cluster can be quantied using the cluster variance, which
is the average of the square of the dierence between the decoded symbol values
(the soft decisions) in z and the nearest member of the alphabet (the nal hard
decisions).
The Matlab function quantalph.m is used in idsys.m to calculate the hard
decisions, which are then converted back into a text character string using
pam2letters.m. If all goes well, this reproduces the original message. The only
aw is that the last symbol of the message has been lost due to the inherent delay
of the lowpass ltering and the pulse-shape correlation. Because four symbols are
needed to decode a single character, the loss of the last symbol also results in
the loss of the last character. The function pam2letters.m provides a friendly
reminder in the Matlab command window that this has happened.
The problems that follow give a few more ways to explore the behavior of the
ideal system.
Exercise 9.1. Using idsys.m, examine the eect of using dierent carrier frequencies. Try fc = 50, 30, 3, 1, 0.5. What are the limiting factors that cause
some to work and others to fail?

Stu Happens

175

Exercise 9.2. Using idsys.m, examine the eect of using dierent oversampling
frequencies. Try M = 1000, 25, 10. What are the limiting factors that cause some
to work and others to fail?
Exercise 9.3. What happens if the LPF at the beginning of the receiver is
removed? What do you think will happen if there are other users present? Try
adding in another user at fc = 30.
Exercise 9.4. What are the limits to the LPF design at the beginning of the
receiver? What is the lowest cuto frequency that works? The highest?
Exercise 9.5. Using the same specications (fbe=[0 0.1 0.2 1]; damps= [1
1 0 0 ];), how short can you make the LPF? Explain.

9.3

Flat Fading: A Simple Impairment and a Simple Fix

Unfortunately, a number of the assumptions made in the simulation of the ideal
system idsys.m are routinely violated in practice. The designer of a receiver
must somehow compensate by improving the receiver. This section presents an
impairment (at fading) for which we have already developed a x (an AGC).
Later sections describe misbehavior due to a wider variety of common impairments that we will spend the rest of the book combating.
Flat fading occurs when there are obstacles moving in the path between the
transmitter and receiver or when the transmitter and receiver are moving with
respect to each other. It is most commonly modeled as a time-varying channel
gain that attenuates the received signal. The modier at implies that the
loss in gain is uniform over all frequencies. (In communications jargon, it is
not frequency-selective.) This section begins by studying the loss of performance
caused by a time-varying channel gain (using a modied version of idsys.m) and
then examines the ability of an adaptive element (the automatic gain control,
AGC) to make things right.
In the ideal system of the preceding section, the gain between the transmitter and the receiver was implicitly assumed to be unity. What happens when
this assumption is violated, when at fading is experienced in midmessage? To
examine this question, suppose that the channel gain is unity for the rst 20%
of the transmission, but that for the last 80% it drops by half. This at fade can
easily be studied by inserting the following code between the transmitter and
the receiver parts of idsys.m.
Listing 9.3. idsysmod1.m modication of idsys.m with time-varying fading channel
l r=length ( r ) ;
% length of transmitted s i gnal vector
f p =[ o n e s ( 1 , f l o o r ( 0 . 2 * l r ) ) , . . .
. . . 0 . 5 * o n e s ( 1 , l r f l o o r ( 0 . 2 * l r ) ) ] ; % f l a t f a d i n g p r o f i l e
r=r . * f p ;
% a p p l y p r o f i l e to t r a n s m i t t e d s i g n a l

176

Chapter 9. Stu Happens

3
2
1
0

Figure 9.10 Soft decisions with

uncompensated at fading.

2
3
4

20

40

60

80 100 120 140 160 180 200

The resulting plot of the soft decisions in Figure 9.10 (obtained via
plot([1:length(z)], z,.)) shows the eect of the fade in the latter 80%
of the response. Shrinking the magnitude of the symbols 3 by half puts them
in the decision region for 1, which generates a large number of symbol errors.
Indeed, the recovered message looks nothing like the original.
Section 6.8 has already introduced an adaptive element designed to compensate for at fading: the automatic gain control, which acts to maintain the
power of a signal at a constant known level. Stripping out the AGC code from
agcvsfading.m on page 128 and combining it with the fading channel just discussed creates a simulation in which the fade occurs, but in which the AGC can
actively work to restore the power of the received signal to its desired nominal
value ds 1.
Listing 9.4. idsysmod2.m modication of idsys.m with fading plus automatic gain control
ds=pow ( r ) ;
l r=length ( r ) ;
f p =[ o n e s ( 1 , f l o o r ( 0 . 2 * l r ) ) , . . .
0 . 5 * o n e s ( 1 , l r f l o o r ( 0 . 2 *
r=r . * f p ;
g=zeros ( 1 , l r ) ; g ( 1 ) = 1 ;
nr=zeros ( 1 , l r ) ;
mu= 0 . 0 0 0 3 ;
f o r i =1: l r 1
nr ( i )=g ( i ) * r ( i ) ;
g ( i +1)=g ( i )mu* ( nr ( i )2 ds ) ;
end
r=nr ;

% d e s i r e d a v e r a g e power o f s i g n a l
% length of transmitted s i gnal
l r ) ) ] ; % f l a t fading p r o f i l e
% a p p l y to t r a n s m i t t e d s i g n a l
% i n i t i a l i z e gain
%
%
%
%

stepsize
a d a p t i v e AGC e l e m e n t
AGC o u t p u t
adapt g a i n

% received signal is

still

called r

Inserting this segment into idsys.m (immediately after the time-varying fading
channel modication) results in only a small number of errors that occur right at
the time of the fade. Very quickly, the AGC kicks in to restore the received power.
The resulting plot of the soft decisions (via plot([1:length(z)],z,.)) in
Figure 9.11 shows how quickly after the abrupt fade the soft decisions return

Stu Happens

177

4
3
2
1
Figure 9.11 Soft decisions with an
AGC compensating for an abrupt
at fade.

0
1
2
3
4

20

40

60

80 100 120 140 160 180 200

AGC gain parameter

2.6
2.2
1.8

Figure 9.12 Trajectory of the

1.4

AGC gain parameter as it moves

to compensate for the fade.

1
0

0.4

0.8

1.2
Iteration

1.6

2
x104

to the appropriate sector. (Look for where the larger soft decisions exceed a
magnitude of 2.)
Figure 9.12 plots the trajectory of the AGC gain g as it moves from the
vicinity of unity to the vicinity of 2 (just what is needed to counteract a 50%
fade). Increasing the stepsize mu can speed up this transition, but also increases
the range of variability in the gain as it responds to short periods when the
square of the received signal does not closely match its long-term average.
Exercise 9.6. Another idealized assumption made in idsys.m is that the
receiver knows the start of each frame; that is, it knows where each foursymbol group begins. This is a kind of frame synchronization problem and
was absorbed into the specication of a parameter l which appears in the code
as 0.5*fl+M. With the default settings, l is 125. This problem poses the following
question. What if this is not known, and how can it be xed?
a. Verify, using idsys.m, that the message becomes scrambled if the receiver
is mistaken about the start of each group of four. Add a random number of
4-PAM symbols before the message sequence, but do not tell the receiver

178

Chapter 9. Stu Happens

that you have done so (i.e., do not change l). What value of l would x the
problem? Can l really be known beforehand?
b. Section 8.5 proposed the insertion of a marker sequence as a way to synchronize the frame. Add a seven-symbol marker sequence just prior to the rst
character of the text. In the receiver, implement a correlator that searches for
the known marker. Demonstrate the success of this modication by adding
random symbols at the start of the transmission. Where in the receiver have
you chosen to put the correlation procedure? Why?
c. One quirk of the system (observed in the eye diagram in Figure 9.8) is that
each group of four begins with a negative number. Use this feature (rather
than a separate marker sequence) to create a correlator in the receiver that
can be used to nd the start of the frames.
d. The previous two exercises showed two possible solutions to the framesynchronization problem. Explain the pros and cons of each method, and
argue which is a better solution.

9.4

Other Impairments: More What Ifs

Of course, a fading channel is not the only thing that can go wrong in a telecommunication system. (Think back to the what if questions in the rst section
of this chapter.) This section considers a range of synchronization and interference impairments that violate the assumptions of the idealized system. Though
each impairment is studied separately (i.e., assuming that everything functions
ideally except for the particular impairment of interest), a single program is
written to simulate any of the impairments. The program impsys.m leaves both
the transmitter and the basic operation of the receiver unchanged; the primary
impairments are to the sampled sequence that is delivered to the receiver.
The rest of this chapter conducts a series of experiments dealing with stu
that can happen to the system. Interference is added to the received signal as
additive Gaussian channel noise and as multipath interference. The oscillator at
the transmitter is no longer presumed to be synchronized with the oscillator at
the receiver. The best sample times are no longer presumed to be known exactly
in either phase or period.
Listing 9.5. impsys.m impairments to the receiver
cng=input ( channel noise gain : try 0 , 0.6 or 2 :: ) ;
c d i=input ( channel multipath : 0 for none ,...
1 for mild or 2 for harsh :: ) ;
f o=input ( tranmsitt er mixer freq offset in percent :...
try 0 or 0.01 :: ) ;
po=input ( tranmsitt er mixer phase offset in rad :...
try 0 , 0.7 or 0.9 :: ) ;
t o p e r=input ( baud timing offset as percent of symb period :...
try 0 , 20 or 30 :: ) ;
s o=input ( symbol period offset : try 0 or 1 :: ) ;

Stu Happens

179

% INSERT TRANSMITTER CODE (FROM IDSYS .M) HERE

i f cdi < 0.5 ,
% channel ISI
mc=[1 0 0 ] ;
% d i s to r ti o n f r e e channel
e l s e i f cdi <1.5 ,
% mild m u l t i p a t h c h a n n e l
mc=[1 zeros ( 1 ,M) 0 . 2 8 zeros ( 1 , 2 . 3 *M) 0 . 1 1 ] ;
else
% harsh multipath channel
mc=[1 zeros ( 1 ,M) 0 . 2 8 zeros ( 1 , 1 . 8 *M) 0 . 4 4 ] ;
end
mc=mc/ ( sqrt (mc*mc ) ) ;
% n o r m a l i z e c h a n n e l power
dv=f i l t e r (mc , 1 , r ) ;
% f i l t e r s i g n a l th r o u g h c h a n n e l
nv=dv+cng * ( randn ( s i z e ( dv ) ) ) ;
% add G a u s s i a n c h a n n e l n o i s e
to=f l o o r ( 0 . 0 1 * t o p e r *M) ;
% f r a c t i o n a l period delay
rnv=nv(1+ to :N*M) ;
% d e l a y i n onsymbol d e s i g n a t i o n
r t =(1+ to )/M: 1 /M: length ( nv )/M;
% time v e c t o r with d e l a y e d s t a r t
rM=M+s o ;
% r e c e i v e r sampler timing o f f s e t
% INSERT RECEIVER CODE (FROM IDSYS .M) HERE

The rst few lines of impsys.m prompt the user for parameters that dene the
impairments. The channel noise gain parameter cng is a gain factor associated
with a Gaussian noise that is added to the received signal. The suggested values
of 0, 0.6, and 2 represent no impairment, mild impairment (that only rarely
causes symbol recovery errors), and a harsh impairment (that causes multiple
symbol errors), respectively.
The second prompt selects the multipath interference: none, mild, or harsh.
In the mild and harsh cases, three copies of the transmitted signal are summed
at the receiver, each with a dierent delay and amplitude. This is implemented
by passing the transmitted signal through a lter whose impulse response is
specied by the variable mc. As occurs in practice, the transmission delays are
not necessarily integer multiples of the symbol interval. Each of the multipath
models has its largest tap rst. If the largest path gain were not rst, this could
be interpreted as a delay between the receipt of the rst sample of the rst pulse
of the message and the optimal sampling instant.
The next pair of prompts concern the transmitter and receiver oscillators.
The receiver assumes that the phase of the oscillator at the transmitter is zero
at the time of arrival of the rst sample of the message. In the ideal system,
this assumption was correct. In impsys.m, however, the receiver makes the same
assumption, but it might no longer be correct. Mismatch between the phase of
the oscillator at the transmitter and the phase of the oscillator at the receiver is
an inescapable impairment (unless there is also a separate communication link or
added signal such as an embedded pilot tone that synchronizes the oscillators).
The user is prompted for a carrier phase oset in radians (the variable po) that
is added to the phase of the oscillator at the transmitter, but not at the receiver.
Similarly, the frequencies of the oscillators at the transmitter and receiver may
dier by a small amount. The user species the frequency oset in the variable
fo as a percentage of the carrier frequency. This is used to scale the carrier
frequency of the transmitter, but not of the receiver. This represents a dierence

180

Chapter 9. Stu Happens

between the nominal values used by the receiver and the actual values achieved
by the transmitter.
Just as the receiver oscillator need not be fully synchronized with the transmitter oscillator, the symbol clock at the receiver need not be properly synchronized
with the transmitter symbol period clock. Eectively, the receiver must choose
when to sample the received signal on the basis of its best guess as to the phase
and frequency of the symbol clock at the transmitter. In the ideal case, the delay
between the receipt of the start of the signal and the rst sample time was readily calculated using the parameter l. But l cannot be known in a real system
because the rst sample depends, for instance, on when the receiver is turned
on. Thus, the phase of the symbol clock is unknown at the receiver. This impairment is simulated in impsys.m using the timing oset parameter toper, which
is specied as a percentage of the symbol period. Subsequent samples are taken
at positive integer multiples of the presumed sampling interval. If this interval
is incorrect, then the subsequent sample times will also be incorrect. The nal
impairment is specied by the symbol period oset, which sets the symbol
period at the transmitter to so less than that at the receiver.
Using impsys.m, it is now easy to investigate how each impairment degrades
the performance of the system.

9.4.1

Additive Channel Noise

Whenever the channel noise is greater than half the gap between two adjacent
symbols in the source constellation, a symbol error may occur. For the constellation of 1s and 3s, if a noise sample has magnitude larger than 1, then the
output of the quantizer may be erroneous.
Suppose that a white, broadband noise is added to the transmitted signal. The spectrum of the received signal, which is plotted in Figure 9.13 (via
plotspec(nv,1/rM)), shows a nonzero noise oor compared with the ideal
(noise-free) spectrum in Figure 9.3. A noise gain factor of cng=0.6 leads to a
cluster variance of about 0.02 and no symbol errors. A noise gain of cng=2 leads
to a cluster variance of about 0.2 and results in approximately 2% symbol errors.
When there are 10% symbol errors, the reconstructed text becomes undecipherable (for the particular coding used in letters2pam.m and pam2letters.m).
Thus, as should be expected, the performance of the system degrades as the
noise is increased. It is worthwhile taking a closer look to see exactly what goes
wrong.
The eye diagram for the noisy received signal is shown in Figure 9.14, which
should be compared with the noise-free eye diagram in Figure 9.8. This is plotted
using the following Matlab commands:
u l=f l o o r ( ( length ( x3 ) 124)/(4 * rM ) ) ;
plot ( reshape ( x3 ( 1 2 5 : u l * 4 * rM+124) ,4 * rM , u l ) ) .

Stu Happens

181

Amplitude

Magnitude

20

40

60

80 100 120 140 160 180 200

Seconds

5000
4000
3000
2000
1000
0
50 40 30 20 10 0 10
Frequency

20

30

40

Figure 9.13 When noise is

added, the received signal
appears jittery. The spectrum
has a noticeable noise oor.

50

5
4
3
2

Figure 9.14 The eye diagram

of the received signal x3

repeatedly overlays
four-symbol-wide segments.
The channel noise is not
insignicant.

0
1
2
3
4
5

50

100

150

200

250

300

350

400

It should be clear from the noisy eye diagram that it would be very dicult to
correctly decode the symbols directly from this signal. Fortunately, the correlation lter reduces the noise signicantly, as shown in the eye diagram in Figure
9.15. (This is plotted as before, substituting y for x3.) Comparing Figures 9.14
and 9.15 closely, observe that the whole of the latter is shifted over in time by
about 50 samples. This is the eect of the time delay of the correlator lter,
which is half the length of the lter. Clearly, it is much easier to correctly decode
using y than using x3, though the pulse shapes of Figure 9.15 are still blurred
when compared with the ideal pulse shapes in Figure 9.8.
Exercise 9.7. The correlation lter in impsys.m is a lowpass lter with impulse
response given by the pulse shape p.
a. Plot the frequency response of this lter. What is the bandwidth of this lter?

182

Chapter 9. Stu Happens

4
3
2
1

Figure 9.15 The eye diagram

of the received signal y after

the correlation lter. The
noise is reduced signicantly.

0
1
2
3
4

50

100

150

200

250

300

350

400

b. Design a lowpass lter using firpm that has the same length and the same
bandwidth as the correlation lter.
c. Use your new lter in place of the correlation lter in impsys.m. Has the
performance improved or worsened? Explain in detail what tests you have
used.
No peeking ahead to Chapter 11.

9.4.2

Multipath Interference
The next impairment is interference caused by a multipath channel, which occurs
whenever there is more than one route between the transmitter and the receiver.
Because these paths experience dierent delays and attenuations, multipath
interference can be modeled as a linear lter. Since lters can have complicated
frequency responses, some frequencies may be attenuated more than others, and
so this is called frequency-selective fading.
The mild multipath interference in impsys.m has three (nonzero) paths
between the transmitter and the receiver. Its frequency response has numerous dips and bumps that vary in magnitude from about +2 to 4 dB. (Verify
this using freqz.) A plot of the soft decisions is shown in Figure 9.16 (from
plot([1:length(z)],z,.)), which should be compared with the ideal constellation diagram in Figure 9.9. The eect of the mild multipath interference
is to smear the lines into stripes. As long as the stripes remain separated, the
quantizer is able to recover the symbols, and hence the message, without errors.
The harsh multipath channel in impsys.m also has three paths between the
transmitter and receiver, but the later reections are larger than in the mild
case. The frequency response of this channel has peaks up to about +4 dB and
down to about 8 dB, so its eects are considerably more severe. The eect
of this channel can be seen directly by looking at the constellation diagram of
the soft decisions in Figure 9.17. The constellation diagram is smeared, and it is

Stu Happens

183

4
3
2
Figure 9.16 With mild multipath
interference, the soft decisions
can readily be segregated into
four stripes that correspond to
the four symbol values.

1
0
1
2
3
4

20

40

60

80

100 120 140 160 180 200

4
3
2
Figure 9.17 With harsher

multipath interference, the soft

decisions smear and it is no
longer possible to see which
points correspond to which of the
four symbol values.

0
1
2
3
4

20

40

60

80

100 120 140 160 180 200

no longer possible to visually distinguish the four stripes that represent the four
symbol values. It is no surprise that the message becomes garbled. As the output
shows, there are about 10% symbol errors, and most of the recovered characters
are wrong.

9.4.3

Carrier Phase Oset

For the receiver in Figure 9.1, the dierence between the phase of the modulating
sinusoid at the transmitter and the phase of the demodulating sinusoid at the
receiver is the carrier phase oset. The eect of a nonzero oset is to scale the
received signal by a factor equal to the cosine of the oset, as was shown in (5.4)
of Section 5.2. Once the phase oset is large enough, the demodulated signal
contracts so that its maximum magnitude is less than 2. When this happens, the
quantizer always produces a 1. Symbol errors abound.

184

Chapter 9. Stu Happens

2
1.5
1
Figure 9.18 Soft decisions for
harsh carrier phase oset are
never greater than 2. The
quantizer nds no 3s and
many symbol errors occur.

0.5
0
0.5
1
1.5
2

20

40

60

80

100 120 140 160 180 200

When running impsys.m, there are two suggested nonzero choices for the
phase-oset parameter po. With po=0.9, cos(0.9) = 0.62, and 3 cos(0.9) < 2.
This is shown in the plot of the soft-decision errors in Figure 9.18. For the
milder carrier phase oset (po=0.7), the soft decisions result in no symbol errors,
because the quantizer will still decode values at 3 cos(0.7) = 2.3 as 3.
As long as the constellation diagram retains distinct horizontal stripes, all is
not lost. In Figure 9.18, even though the maximum magnitude is less than 2,
there are still four distinct stripes. If the quantizer could be scaled properly, the
symbols could be decoded successfully. Such a scaling might be accomplished,
for instance, by another AGC, but such scaling would not improve the signalto-noise ratio. A better approach is to identify the unknown phase oset, as
discussed in Chapter 10.
Exercise 9.8. Using impsys.m as a basis, implement an AGC-style adaptive
element to compensate for a phase oset. Verify that your method works for a
phase oset of 0.9 and for a phase oset of 1.2. Show that the method fails when
the phase oset is /2.

9.4.4

Carrier Frequency Oset

The receiver in Figure 9.1 has a carrier frequency oset when the frequency of
the carrier at the transmitter diers from the assumed frequency of the carrier
at the receiver. As was shown in (5.5) in Section 5.2, this impairment is like
a modulation by a sinusoid with frequency equal to the oset. This modulating eect is catastrophic when the low-frequency modulator approaches a zero
crossing, since then the gain of the signal approaches zero. This eect is apparent
for a 0.01% frequency oset in impsys.m in the plot of the soft decisions (via
plot([1:length(z)],z,.)) in Figure 9.19. This experiment suggests that the
receiver mixer frequency must be adjusted to track that of the transmitter.

Stu Happens

185

3
2
1
Figure 9.19 Soft decisions for
0.01% carrier frequency oset.

0
1
2
3

9.4.5

20

40

60

80

100 120 140 160 180 200

Downsampler Timing Oset

As shown in Figure 9.7, there is a sequence of best times at which to downsample. When the starting point is correct and no intersymbol interference (ISI)
is present, as in the ideal system, the sample times occur at the top of the pulses.
When the starting point is incorrect, all the sample times are shifted away from
the top of the pulses. This was set in the ideal simulation using the parameter
l, with its default value of 125. The timing-oset parameter toper in impsys.m
is used to oset the received signal. Essentially, this means that the best value
of l has changed, though the receiver does not know it.
This is easiest to see by drawing the eye diagram. Figure 9.20 shows an overlay
of four-symbol wide segments of the received signal (using the reshape command
as in the code on page 173). The receiver still thinks the best times to sample
are at l + nT , but this is clearly no longer true. In fact, whenever the sample
time begins between 100 and 140 (and lies in this or any other shaded region),
there will be errors when quantizing. For example, all samples taken at 125 lie
between 1, and hence no symbols will ever be decoded at their 3 value. In
fact, some even have the wrong sign! This is a far worse situation than in the
carrier phase impairment because no simple amplitude scaling will help. Rather,
a solution must correct the problem; it must slide the times so that they fall
in the unshaded regions. Because these unshaded regions are wide open, this is
often called the open-eye region. The goal of an adaptive element designed to x
the timing-oset problem is to open the eye as wide as possible.

9.4.6

Downsampler Period Oset

When the assumed period of the downsampler is in error, there is no hope. As
mentioned in the previous impairment, the receiver believes that the best times

186

Chapter 9. Stu Happens

Assumed "best times" to take samples

3
2
Figure 9.20 Eye diagram with
downsampler timing oset of
50%. Sample times used by the
ideal system are no longer valid,
and lead to numerous symbol
errors.

1
0
1
2
3

50

100

150

200

250

300

350

400

to sample are at l + nT . When there is a period oset, it means that the value
of T used at the receiver diers from the value actually used at the transmitter.
The prompt in impsys.m for symbol period oset suggests trying 0 or 1. A
response of 1 results in the transmitter creating the signal assuming that there
are M 1 samples per symbol period, while the receiver retains the setting of
M samples per symbol, which is used to specify the correlator lter and to
pick subsequent downsampling instants once the initial sample time has been
selected. The symptom of a misaligned sample period is a periodic collapse of
the constellation, similar to that observed when there is a carrier frequency oset
(recall Figure 9.19). For an oset of 1, the soft decisions are plotted in Figure
9.21. Can you connect the value of the period of this periodic collapse to the
parameters of the simulated example?

9.4.7

Repairing Impairments
When stu happens and the receiver continues to operate as if all were well,
the transmitted message can become unintelligible. The various impairments of
the preceding sections point the way to the next step in the design process by
showing the kinds of problems that may arise. Clearly, the receiver must be
improved to counteract these impairments.
Coding (Chapter 14) and matched receive ltering (Chapter 11) are intended
primarily to counter the eects of noise. Equalization (Chapter 13) compensates
for multipath interference, and can reject narrowband interferers. Carrier recovery (Chapter 10) will be used to adjust the phase, and possibly the frequency
as well, of the receiver oscillator. Timing recovery (Chapter 12) aims to reduce
downsampler timing and period oset. All of these xes can be viewed as digital
signal processing (DSP) solutions to the impairments explored in impsys.m.

Stu Happens

3
2
1
0
1
2
3
3
2
1
0
1
2
3

187

Figure 9.21 When there is a

50

20

100

40

150

60

80

200

250

100 120

300

140

160

350

180

400

1% downsampler period
oset, all is lost, as shown
by the eye diagram in the
top plot and the soft
decisions in the bottom.

200

Each of these xes will be designed separately, as if the problem it is intended

to counter were the only problem in the world. Fortunately, somehow they can
all work together simultaneously. Examining possible interactions between the
various xes, which is normally a part of the testing phase of a receiver design,
will be part of the receiver design project of Chapter 15.

9.5

A B 3IG Deal
While the idealized communication systems in idsys.m and impsys.m illustrate
the basic operation of a transmitterreceiver pair, practical systems are a bit
more complicated. One of the problems is keeping track of all the problems that
can arise! In order to help organize all the possibilities, this section introduces
the Basic Black Box Impairment Generator, or B 3IG, as shown in Figure 9.22.

MATLAB
script
defining
transmitter
parameters

Signal r as it
enters the receiver
Basic Black Box
Impairment Generator

Figure 9.22 The Basic Black Box Impairment Generator, or B 3IG, inputs a Matlab

script that species the operation of the transmitter along with any desired
impairments. The output is the (received) signal r that enters the receiver. You can
nd out how the system operates by peering ahead to Appendix H, but we really
shouldnt mention it at this point.

188

Chapter 9. Stu Happens

The B 3IG transmitter is a souped-up version of the idealized transmitters in

idsys.m and impsys.m. It allows easy incorporation of a number of impairments,
singly or in combination, and is especially useful when examining possible interactions between the various receiver strategies and xes that are discussed in
forthcoming chapters. Though a complete description of the B 3IG transmitter is
presented in Appendix H, it is probably not useful to jump ahead at this point.
Better to think of B 3IG as a tool for exploration, for generating noisy received
signals with various user-specied and user-controlled impairments. In keeping
with the theme of this chapter, there is nothing genuinely new in B 3IG; the basic
building blocks are the same as those in the previous chapters.
Using the B 3IG transmitter is easy: just type the command in Listing 9.6.
Listing 9.6. This code calls the B 3IG transmitter. The received signal is in r and the
transmitted (encoded) symbols are in s.
[ r , s ]= B i g T r a n s m i t t e r (m, frameParams , . . .
rfParams , chanParams ) ;

There are four inputs to the B 3IG. The message that is to be transmitted is specied in the variable m. The frame parameters frameParams include a specication
for how many symbols are included in each block and a string that represents
the preamble (or header). These are familiar from Chapter 8. In order to mimic
the idealized transmission in idsys.m and impsys.m, the length of a block is set
to 1 and there is no preamble. The message includes the Oscar Meyer jingle and
some numbers. The complete list of parameters is given in Listing 9.7.
Listing 9.7. Parameters in the script BigIdeal.m allow the B 3IG transmission to mimic
the idealized system idsys.m and impsys.m.
% Message Gen er a ti o n
m= 01234 I wish I were an Oscar Meyer ...
wiener 56789 ;
% Frame p a r a m e te r s
frameParams . u s er D a ta L en g th =1;
frameParams . preamble= ;
frameParams . chanCodingFlag =0;
frameParams . b i t E n c o d i n g F l a g =0;
% Channel p a r a m eter s , Adj Users , I n t . ,
chanParams . c1 =[1 0 0 ] ;
chanParams . c2 =[1 0 0 ] ;
chanParams . randomWalkVariance=0;
chanParams .SNR=I n f ;
chanParams . a d j a c e n t U s er 1 P o w er=I n f ;
chanParams . a d j a c e n t U s e r 1 f i f =0;
chanParams . a d j a c e n t U s er 1 C h a n =[1 0 0 ] ;
chanParams . a d j a c e n t U s er 2 P o w er=I n f ;
chanParams . a d j a c e n t U s e r 2 f i f =0;
chanParams . a d j a c e n t U s er 2 C h a n =[1 0 0 ] ;
chanParams . NBIfreq =0;
chanParams . NBIPower=I n f ;
% RF P a r a m e t e r s

Stu Happens

189

rfParams . f i f e r r =0;
rfParams . T t e r r =0;
rfParams . p h a s e N o i s e V a r i a n c e =0;
rfParams . SRRCLength =4;
rfParams . S R R C r o l l o f f = 0 . 3 ;
rfParams . f s =100;
rfParams . T t =1;
rfParams . f i f =20;

The channel parameters chanParams specify how the message is distorted as

it moves from the transmitter to the receiver. Intersymbol interference, additive
noise, and interference from other users can all be modeled by changing these
values. Finally, the RF parameters rfParams specify the parameters used: the
frequency of the transmitter, the symbol rate, the size of the pulse shape, the
intermediate frequency, etc. All frequency values are specied in Hz, and all time
values are given in seconds. As before, the values in BigIdeal.m of Listing 9.7
mimic the idealized transmission.
Exercise 9.9. One dierence between BigIdeal.m and the received signal in
impsys.m is that the pulse shape is a square-root raised cosine pulse instead of
a Hamming pulse. Run the script and plot the spectrum of the received signal
in BigIdeal.m. Compare this with the spectrum of the received signal from
impsys.m. How do the spectra change when the SNR of the additive channel
noise is decreased?
Exercise 9.10. A rudimentary receiver is provided in Listing 9.8. Observe that
the code makes errors when noise is added because the gain (the parameter
receiver gain) has not been chosen appropriately. Find a new choice of gain
that permits the receiver to operate error-free even with the level of noise present.
How did you nd this choice of gain?
Listing 9.8. BigIdeal.m (part 2) idealized B 3IG receiver
% A d j u s t a b l e r e c e i v e r p a r a m eter s
r e c e i v e r g a i n =1;
d a t a s t a r t =5;
% Fixed r e c e i v e r p a r a m e te r s
M=100;
% upsampling r a t i o
f c =20;
% c a r r i e r frequency
% l e n g t h o f p u l s e shape
s r r c l e n g t h =4;
% am d em o d u l a ti o n o f r e c e i v e d s i g n a l s e q u e n c e r
c2=cos (2 * pi * f c * [ 1 /M: 1 /M: length ( r )/M] ) ; % c o s i n e f o r mixing
x2=r . * c2 ;
% demod r e c e i v e d s i g n a l
% c o r r e l a t e ( f i l t e r ) to e x t r a c t upsampled p u l s e s
y=f i l t e r ( s r r c ( s r r c l e n g t h , 0 . 3 , 1 0 0 ) , 1 , x2 ) ; % f i l t e r with p u l s e
% s e t d e l a y to f i r s t symbolsample and i n c r e m e n t by M
% downsample to symbol r a t e
z=y ( s r r c l e n g t h *M:M: end ) ;
% t o s s out f i r s t few symbols , s c a l e by g a i n
z=z ( d a t a s t a r t : d a t a s t a r t+length (m) * 4 1) * r e c e i v e r g a i n ;
plot ( [ 1 : length ( z ) ] , z , . )
% plot soft decisions
% d e c i s i o n d e v i c e and symbol matching p e r f o r m a n c e a s s e s s m e n t

190

Chapter 9. Stu Happens

mprime=q u a n t a l p h ( z , [ 3 , 1 , 1 , 3 ] ) ;
% q u a n t i z e to a l p h a b e t
c v a r =(mprimez ) * ( mprimez ) / length ( mprime ) % c l u s t e r v a r i a n c e
lmp=length ( mprime ) ;
p e r e r r =100 *sum( abs ( sign ( mprimel e t t e r s 2 p a m (m) ) ) ) / lmp % symb e r r
% decode d e c i s i o n d ev i ce output to tex t s t r i n g
r e c o n s t r u c t e d m e s s a g e=p a m 2 l e t t e r s ( mprime )

Exercise 9.11. Use the AGC code from agcgrad.m to nd the proper gain for
the receiver in BigIdeal.m. If the AGC converges slowly, you may wish to use a
longer message m.
Exercise 9.12. Observe that when the parameter dstart is specied incorrectly,
the message is not reconstructed correctly, even when there is no additive noise.
Implement a strategy to identify dstart from the received signal. Hint: perhaps
a header or frame would help.
Exercise 9.13. How would you change the frame, channel, or RF parameters
to reproduce Figures 9.16 and 9.17?
Exercise 9.14. How would you change the frame, channel, or RF parameters
to reproduce Figure 9.19?

Step 4: The Adaptive Components

The fourth step describes all the practical xes that are needed in order to create
a workable radio. One by one the various pragmatic problems are studied and
solutions are proposed, implemented, and tested. These include xes for additive
noise, for timing oset problems, for clock frequency mismatches and jitter, and
for multipath reections. The order in which topics are discussed is the order in
which they appear in the receiver.
Carrier recovery
the timing of frequency translation
Receive ltering
the design of pulse shapes
Clock recovery
the timing of sampling
Equalization
lters that adapt to the channel
Coding
making data resilient to noise

Chapter 10
Chapter 11
Chapter 12
Chapter 13
Chapter 14

10 Carrier Recovery

Figure 10.1 shows a generic transmitter and receiver pair that emphasizes the
modulation and corresponding demodulation. Even assuming that the transmission path is ideal (as in Figure 10.1), the signal that arrives at the receiver is a
complicated analog waveform that must be downconverted and sampled before
the message can be recovered. For the demodulation to be successful, the receiver
must be able to gure out both the frequency and the phase of the modulating
sinusoid used in the transmitter, as was shown in (5.4) and (5.5) and graphically
illustrated in Figures 9.18 and 9.19. This chapter discusses a variety of strategies
that can be used to estimate the phase and frequency of the carrier and to x the
gain problem (of (5.4) and Figure 9.18) and the problem of vanishing amplitudes
(in (5.5) and Figure 9.19). This process of estimating the frequency and phase
of the carrier is called carrier recovery.
Figure 10.1 shows two downconversion steps: one analog and one digital. In
a purely analog system, no sampler or digital downconversion would be needed.
The problem is that accurate analog downconversion requires highly precise analog components, which can be expensive. In a purely digital receiver, the sampler
would directly digitize the received signal, and no analog downconversion would
be required. The problem is that sampling this fast can be prohibitively expensive. The happy compromise is to use an inexpensive analog downconverter to
translate to some lower intermediate frequency, at which it is possible to sample
cheaply enough. At the same time, sophisticated digital processing can be used to
compensate for inaccuracies in the cheap analog components. Indeed, the same
adaptive elements that estimate and remove the unknown phase oset between
the transmitter and the receiver automatically compensate for any additional
phase inaccuracies in the analog portion of the receiver.
Normally, the transmitter and receiver agree to use a particular frequency for
the carrier, and in an ideal world, the frequency of the carrier of the transmitted
signal would be known exactly. But even expensive oscillators may drift apart
in frequency over time, and cheap (inaccurate) oscillators may be an economic
necessity. Thus, there needs to be a way to align the frequency of the oscillator at

Carrier Recovery

Message
m(kT)

Received
signal
Pulse
shaping

Modulation
fc

Analog
conversion
to IF

Ts

Digital downconversion
to baseband

Resampling

193

Reconstructed
message
m(kT)
Decision

Timing

synchronization
Frequency and

phase tracking
f0

Figure 10.1 Diagram of a communications system emphasizing the need for

synchronization with the frequency fc and phase of the carrier. The
frequency-tracking element adjusts the parameter f0 in an attempt to match fc and
the phase-tracking element adjusts the parameter in an attempt to match .

the transmitter with the frequency of the oscillator at the receiver. Since the goal
is to nd the frequency and phase of a signal, why not use a Fourier transform
(or, more properly, an FFT)? Section 10.1 shows how to isolate a sinusoid that is
at twice the frequency of the carrier by squaring and ltering the received signal.
The frequency and phase of this sinusoid can then be found in a straightforward
manner by using the FFT, and the frequency and phase of the carrier can then
be simply deduced. Though feasible, this method is rarely used because of the
computational cost.
The strategy of the following sections is to replace the FFT operation with
an adaptive element that achieves its optimum value when the phase of an estimated carrier equals the phase of the actual carrier. By moving the estimates in
the direction of the gradient of a suitable performance function, the element can
recursively hone in on the correct value. Assuming initially that the frequency
is known, there are various ways to structure adaptive elements that iteratively
estimate the unknown phase of a carrier. One such performance function, discussed in Section 10.2, is the square of the dierence between the received signal
and a locally generated sinusoid. Another performance function leads to the
well-known phase-locked loop, which is discussed in depth in Section 10.3, and
yet another performance function leads to the Costas loop of Section 10.4. An
alternative approach uses the decision-directed method detailed in Section 10.5.
Each of these methods is derived from an appropriate performance function, each
is simulated in Matlab, and each can be understood by looking at the appropriate error surface. This approach should be familiar from Chapter 6, where it
was used in the design of the AGC.
Section 10.6 then shows how to modify the adaptive elements to attack the
frequency-estimation problem. Three ways are shown. The rst tries (unsuccessfully) to apply a direct adaptive method, and the reasons for the failure
provide a cautionary counterpoint to the indiscriminate application of adaptive
elements. The second, a simple indirect method detailed in Section 10.6.2, uses
two loops. Since the phase of a sinusoid is the derivative of its frequency, the
rst loop tracks a line (the frequency oset) and the second loop ne-tunes
the estimation of the phase. The third technique, in Section 10.6.3, shows how

194

Chapter 10. Carrier Recovery

the dual-loop method can be simplied and generalized by using an integrator

within a single phase-locked loop. This forms the basis for an eective adaptive
frequency-tracking element.

10.1

Phase and Frequency Estimation via an FFT

As indicated in Figure 10.1, the received signal consists of a message m(kT )
modulated by the carrier. In the simplest case, when the modulation uses AM
with a large carrier as in Section 5.1, it may be quite easy to locate the carrier
and its phase. More generally, however, the carrier will be well hidden within the
received signal and some kind of extra processing will be needed to bring it to
the foreground.
To see the nature of the carrier-recovery problem explicitly, the following code
generates two dierent received signals: the rst is AM modulated with large
carrier and the second is AM modulated with suppressed carrier. The phase
and frequencies of both signals can be recovered using an FFT, though the
suppressed-carrier scheme requires additional processing before the FFT can
successfully be applied.
Drawing on the code in pulseshape.m on page 156, and modulating with
the carrier c, pulrecsig.m creates the two dierent received signals. The pam
command creates a random sequence of symbols drawn from the alphabet 1,
3, and then uses hamming to create a pulse shape.1 The oversampling factor M
is used to simulate the analog portion of the transmission, and M Ts is equal
to the symbol time T .
Listing 10.1. pulrecsig.m make pulse-shaped signal
N=10000; M=20; Ts = . 0 0 0 1 ;
time=Ts *N*M; t=Ts : Ts : time ;
m=pam (N, 4 , 5 ) ;
mup=zeros ( 1 ,N*M) ;
mup ( 1 :M:N*M)=m;
ps=hamming (M) ;
s=f i l t e r ( ps , 1 , mup ) ;
f c =1000; p h o f f = 1.0;
c=cos (2 * pi * f c * t+p h o f f ) ;
r s c=s . * c ;
r l c =( s + 1 ). * c ;

% # symbols , o v e r s a m p l i n g f a c t o r
% s a m p l i n g i n t e r v a l & time v e c t o r
% 4 l e v e l s i g n a l o f l e n g t h N
%
%
%
%
%
%
%

o v er s a m p l e by i n t e g e r l e n g t h M
b l i p p u l s e o f width M
c o n v o l v e p u l s e shape with data
c a r r i e r f r e q and phase
construct c a r r i e r
modulated s i g n a l ( s m a l l c a r r i e r )
modulated s i g n a l ( l a r g e c a r r i e r )

Figure 10.2 plots the spectra of the large- and suppressed-carrier signals rlc
and rsc. The carrier itself is clearly visible in the top plot, and its frequency and
phase can readily be found by locating the maximum value in the FFT:
f f t r l c =f f t ( r l c ) ;
[m, imax]=max( abs ( f f t r l c ( 1 : end / 2 ) ) ) ;
s s f =(0: length ( t ) 1)/( Ts * length ( t ) ) ;
1

% spectrum o f r l c
% i n d e x o f max peak
% frequency vector

This is not a common (or a particularly useful) pulse shape. It is just easy to use. Good
pulse shapes are considered in detail in Chapter 11.

Carrier Recovery

1000

2000

3000

4000

5000

1000

2000

3000

4000

5000

1000

2000

3000

4000

5000

195

Figure 10.2 The magnitude spectrum of the received signal of a system using AM with

large carrier has a prominent spike at the frequency of the carrier, as shown in the top
plot. When using the suppressed-carrier method in the middle plot, the carrier is not
clearly visible. After preprocessing of the suppressed-carrier signal using the scheme
in Figure 10.3, a spike is clearly visible at twice the desired frequency (and with twice
the desired phase). In the time domain, this corresponds to an undulating sine wave.
f r e q L=s s f ( imax )
phaseL=angle ( f f t r l c ( imax ) )

% f r e q a t th e peak
% phase a t th e peak

Changing the default phase oset phoff changes the phaseL variable accordingly.
Changing the frequency fc of the carrier changes the frequency freqL at which
the maximum occurs. Note that the max function used in this fashion returns
both the maximum value m and the index imax at which the maximum occurs.
On the other hand, applying the same code to the FFT of the suppressedcarrier signal does not recover the phase oset. In fact, the maximum often
occurs at frequencies other than the carrier, and the phase values reported bear
no resemblance to the desired phase oset phoff. There needs to be a way to
process the received signal to emphasize the carrier.
A common scheme uses a squaring nonlinearity followed by a bandpass lter,
as shown in Figure 10.3. When the received signal r(t) consists of the pulsemodulated data signal s(t) times the carrier cos(2f c t + ), the output of the
squaring block is
r2 (t) = s2 (t) cos2 (2fc t + ).

(10.1)

This can be rewritten using the identity 2 cos2 (x) = 1 + cos(2x) in (A.4) to produce
r2 (t) = (1/2)s2 (t)[1 + cos(4fc t + 2)].
Rewriting s2 (t) as the sum of its (positive) average value and the variation about
this average yields
s2 (t) = s2avg + v(t).

196

Chapter 10. Carrier Recovery

Thus,
r2 (t) = (1/2)[s2avg + v(t) + s2avg cos(4fc t + 2) + v(t) cos(4fc t + 2)].
A narrow bandpass lter centered near 2fc passes the pure cosine term in r2 and
suppresses the DC component, the (presumably) lowpass v(t), and the upconverted v(t). The output of the bandpass lter is approximately
1 2
s cos(4fc t + 2 + ),
(10.2)
2 avg
where is the phase shift added by the BPF at frequency 2fc . Since is known
at the receiver, rp (t) can be used to nd the frequency and phase of the carrier. Of
course, the primary component in rp (t) is at twice the frequency of the carrier,
the phase is twice the original unknown phase, and it is necessary to take
into account. Thus some extra bookkeeping is needed. The amplitude of rp (t)
undulates slowly as s2avg changes.
The following Matlab code carries out the preprocessing of Figure 10.3. First,
run pulrecsig.m to generate the suppressed carrier signal rsc.
rp (t) = BPF{r2 (t)}

Listing 10.2. pllpreprocess.m send received signal through square and BPF
r=r s c ;
q=r . 2 ;
f l =500; f f =[0 . 3 8 . 3 9 . 4 1 . 4 2 1 ] ;
f a =[0 0 1 1 0 0 ] ;
h=f i r p m ( f l , f f , f a ) ;
rp=f i l t e r ( h , 1 , q ) ;

%
%
%
%
%
%

suppressed c a r r i e r r
square n o n l i n e a r i t y
BPF c e n t e r f r e q u e n c y a t . 4
which i s t w i c e f c
BPF d e s i g n v i a f i r p m
f i l t e r gives preprocessed r

Then the phase and frequency of rp can be found directly by using the FFT.
% r e c o v e r unknown f r e q and phase u s i n g
f f tr B P F=f f t ( rp ) ;
%
[m, imax]=max( abs ( f f tr B P F ( 1 : end / 2 ) ) ) ; %
s s f =(0: length ( rp ) ) / ( Ts * length ( rp ) ) ; %
f r e q S=s s f ( imax )
%
phasep=angle ( f f tr B P F ( imax ) ) ;
%
[ IR , f ]= f r e q z ( h , 1 , length ( rp ) , 1 / Ts ) ;
%
[ mi , im]=min( abs ( f f r e q S ) ) ;
%
phaseBPF=angle ( IR ( im ) ) ;
%
phaseS=mod ( phasep phaseBPF , pi )
%

FFT
spectrum o f rBPF
f i n d f r e q o f max peak
frequency vector
f r e q a t th e peak
phase a t th e peak
freq response of f i l t e r
f r e q where peak o c c u r s
< o f BPF a t peak f r e q
estimated angle

Observe that both freqS and phaseS are twice the nominal values of fc and
phoff, though there may be a ambiguity (as will occur in any phase estimation).
The intent of this section is to clearly depict the problem of recovering the frequency and phase of the carrier even when it is buried within the data-modulated
signal. The method used to solve the problem (application of the FFT) is not
common, primarily because of the numerical complexity. Most practical receivers
use some kind of adaptive element to iteratively locate and track the frequency
and phase of the carrier. Such elements are explored in the remainder of this
chapter.

Carrier Recovery

r(t)

X2
Squaring
nonlinearity

r2(t)
BPF

197

rp(t) ~ cos(4 fc t + 2 + )

Center frequency
~ 2fc

Figure 10.3 Preprocessing the input to a PLL via a squaring nonlinearity and BPF
results in a sinusoidal signal at twice the frequency with a phase oset equal to twice
the original plus a term introduced by the known bandpass ltering.

Exercise 10.1. The squaring nonlinearity is only one possibility in the

pllpreprocess.m routine.
a. Try replacing the r2 (t) with |r(t)|. Does this result in a viable method of
emphasizing the carrier?
b. Try replacing the r2 (t) with r3 (t). Does this result in a viable method of
emphasizing the carrier?
c. Can you think of other functions that will result in viable methods of emphasizing the carrier?
d. Will a linear function work? Why or why not?
Exercise 10.2. Determine the phase shift at the frequency 2fc caused by the
BPF when
a. fl=490, 496, 502,
b. Ts=0.0001, 0.000101,
c. M=19, 20, 21. Explain why should depend on fl, Ts, and M.

10.2

Squared Dierence Loop

The problem of phase tracking is to determine the phase of the carrier and
to follow any changes in using only the received signal. The frequency fc of
the carrier is assumed known, though ultimately it too must be estimated. The
received signal can be preprocessed (as in the previous section) to create a signal
that strips away the data, in essence fabricating a sinusoid which has twice the
frequency at twice the phase of the unmodulated carrier. This can be idealized
to
rp (t) = cos(4fc t + 2),

(10.3)

which suppresses2 the dependence on the known phase shift of the BPF and
sets the constant s2avg /2 to unity (compare with (10.2)). The form of rp (t) implies
that there is an essential ambiguity in the phase since can be replaced by
+ n for any integer n without changing the value of (10.3). What can be
done to recover (modulo ) from rp (t)?
2

An example that takes into account is given in Exercise 10.8.

198

Chapter 10. Carrier Recovery

Is there some way to use an adaptive element? Section 6.5 suggested that there
are three steps to the creation of a good adaptive element: setting a goal, nding
a method, and then testing. As a rst try, consider the goal of minimizing the
average of the squared dierence between rp (t) and a sinusoid generated, using
an estimate of the phase; that is, seek to minimize
1
(10.4)
avg{(rp (kTs ) cos(4f0 kTs + 2))2 }
4
by choice of , where rp (kTs ) is the value of rp (t) sampled at time kTs and
where f0 is presumed equal to fc . (The subscript SD stands for squared dierence, and is used to distinguish this performance function from others that will
appear in this and other chapters.) This goal makes sense because, if could be
found so that = + n, then the value of the performance function would be
zero. When = + n, then rp (kTs ) = cos(4f0 kTs + 2), e(, k) = 0, and so
JSD () > 0. Hence, (10.4) is minimized when has correctly identied the phase
oset, modulo the inevitable ambiguity.
While there are many methods of minimizing (10.4), an adaptive element that
descends the gradient of the performance function JSD () leads to the algorithm3

dJSD () 
[k + 1] = [k]
,
(10.5)
d =[k]
JSD () = avg{e2 (, k)} =

which is the same as (6.5) with the variable changed from x to . Using the
approximation detailed in (G.12), which holds for small , the derivative and
the average commute. Thus,
 2

d avg{e2 (, k)}
de (, k)
dJSD ()
=
avg
d
d
d


1
de(, k)
= avg e(, k)
(10.6)
2
d
= avg{(rp (kTs ) cos(4f0 kTs + 2)) sin(4f0 kTs + 2)}.
Substituting this into (10.5) and evaluating at = [k] gives4
[k + 1] = [k] avg{(rp (kTs ) cos(4f0 kTs + 2[k]))
sin(4f0 kTs + 2[k])}.

(10.7)

This is implemented in pllsd.m for a phase oset of phoff=-0.8 (i.e., of (10.3)

is 0.8, though this value is unknown to the algorithm). Figure 10.4 plots the
estimates theta for 50 dierent initial guesses theta(1). Observe that many
converge to the correct value at 0.8. Others converge to 0.8 + (about 2.3)
and to 0.8 (about 4).
Listing 10.3. pllsd.m phase tracking minimizing SD
Ts =1/10000; time =1; t =0: Ts : time Ts ;
3
4

% time i n t e r v a l & v e c t o r

Recall the discussion surrounding the AGC elements in Chapter 6.

Recall the convention that [k] = (kTs ) = (t)|t=kTs .

Carrier Recovery

199

Phase offset

2
1

Figure 10.4 The phase-tracking

algorithm (10.7) converges to
the correct phase oset (in this
case 0.8 or to some multiple
0.8 + n) depending on the
initial estimate.

0
1
2
3
0

5
6
Time

10

f c =100; p h o f f = 0.8;
% c a r r i e r f r e q and phase
rp=cos (4 * pi * f c * t +2* p h o f f ) ;
% si m p l i f i e d rec s i gnal
mu= . 0 0 1 ;
% algorithm s t e p s i z e
t h e t a=zeros ( 1 , length ( t ) ) ; t h e t a ( 1 ) = 0 ;
% i n i t i a l i z e estimates
f l =25; h=o n e s ( 1 , f l )/ f l ;
% averaging c o e f f i c i e n t s
z=zeros ( 1 , f l ) ; f 0=f c ;
% b u f f e r f o r avg
f o r k =1: length ( t )1
% run a l g o r i t h m
f i l t i n =(rp ( k)cos (4 * pi * f 0 * t ( k)+2 * t h e t a ( k ) ) ) . . .
* s i n (4 * pi * f 0 * t ( k)+2 * t h e t a ( k ) ) ;
z =[ z ( 2 : f l ) , f i l t i n ] ;
% z contains past inputs
t h e t a ( k+1)= t h e t a ( k)mu* f l i p l r ( h ) * z ;
% update = z c o n v o l v e h
end

Observe that the averaging (a kind of lowpass lter, as discussed in Appendix

G) is not implemented using the filter or conv commands because the complete
input is not available at the start of the simulation. Instead, the time-domain
method is used, and the code here may be compared to the fourth method in
waystofilt.m on page 143. At each time k, there is a vector z of past inputs.
These are multiplied, point by point, with the impulse response h, which is ipped
in time so that the sum properly implements a convolution. Because the lter is
just a moving average, the impulse response is constant (1/fl) over the length
of the ltering.
Exercise 10.3. Use the preceding code to play with the SD phase-tracking
algorithm.
a.
b.
c.
d.

How does the stepsize mu aect the convergence rate?

What happens if mu is too large (say mu=10)?
Does the convergence speed depend on the value of the phase oset?
How does the nal converged value depend on the initial estimate theta(1)?

Exercise 10.4. Investigate these questions by making suitable modications to

pllsd.m.

200

Chapter 10. Carrier Recovery

a. What happens if the phase slowly changes over time? Consider a slow, smallamplitude undulation in phoff.
b. Consider a slow linear drift in phoff.
c. What happens if the frequency f0 used in the algorithm is (slightly) dierent
from the frequency fc used to construct the carrier?
d. What happens if the frequency f0 used in the algorithm is greatly dierent
from the frequency fc used to construct the carrier?
Exercise 10.5. How much averaging is necessary? Reduce the length of the
averaging lter. Is it possible to make the algorithm work with no averaging?
Why does this work? Hint: yes, it is possible. Consider the relationship between
(10.7) and (G.4).
Exercise 10.6. Derive (10.6), following the technique used in Example G.3.
The performance function JSD () of (10.4) provides a mathematical statement
of the goal of an adaptive phase-tracking element. The method is dened by the
algorithm (10.7) and simulations such as pllsd.m suggest that the algorithm
can function as desired. But why does it work?
One way to understand adaptive elements, as discussed in Section 6.6 and
shown in Figure 6.17 on page 119, is to draw the error surface for the performance function. But it is not immediately clear what this looks like, since
JSD () depends on the frequency f0 , the time kTs , and the unknown (through
rp (kTs )), as well as the estimate . Recognizing that the averaging operation acts
as a kind of lowpass lter (see Appendix G if this makes you nervous) allows
considerable simplication of JSD (). Rewrite (10.4) as
JSD () =

1
LPF{(rp (kTs ) cos(4f0 kTs + 2))2 }.
4

(10.8)

Substituting rp (kTs ) from (10.3) and assuming fc = f0 , this can be rewritten

1
LPF{(cos(4f0 kTs + 2) cos(4f0 kTs + 2))2 }.
4
Expanding the square gives
JSD () =

JSD () =

1
LPF{cos2 (4f0 kTs + 2) 2 cos(4f0 kTs + 2)
4
cos(4f0 kTs + 2) + cos2 (4f0 kTs + 2)}.

Using the trigonometric formula (A.4) for the square of a cosine and the formula
(A.13) for the cosine angle sum (i.e., expand cos(x + y) with x = 4f0 kTs and
y = 2, and then again with y = 2) yields
JSD () =

1
LPF{2 + cos(8f0 kTs + 4) 2 cos(2 2)
8
2 cos(8f0 kTs + 2 + 2) + cos(8f0 kTs + 4)}.

Carrier Recovery

201

JSD

0.8
0.4

Figure 10.5 The error surface (10.9)

for the SD phase-tracking algorithm
is shown in the top plot. Analogous
error surfaces for the phase-locked
loop (10.11) and the Costas loop
(10.13) are shown in the middle and
bottom plots. All have minima (or
maxima) at the desired locations (in
this case 0.8) plus n osets.

JPLL

0
0.4
0

JC

0.4
0.8
0.4
0
4

0
2
4
Phase estimates

Since the LPF is linear,

1
1 1
= + LPF{cos(8f0 kTs + 4)} LPF{cos(2 2)}
4 8
4
1
1
LPF{cos(8f0 kTs + 2 + 2)} + LPF{cos(8f0 kTs + 4)}.
4
8
Assuming that the cuto frequency of the lowpass lter is less than 4f0 , this
simplies to
JSD ()

1
(1 cos(2 2)),
4

(10.9)

which is shown in the top plot of Figure 10.5 for = 0.8. The algorithm (10.7)
is initialized with [0] at some point on the surface of the undulating sinusoidal
curve. At each iteration of the algorithm, it moves downhill. Eventually, it will
reach one of the nearby minima, which occur at = 0.8 n for some n.
Thus, Figure 10.5 provides evidence that the algorithm can successfully locate
the unknown phase, assuming that the preprocessed signal rp (t) has the form
of (10.3).
Figure 10.6 shows the algorithm (10.7) with the averaging operation replaced
by the more general LPF. In fact, this provides a concrete answer to Exercise 10.5;
the averaging, the LPF, and the integral block all act as lowpass lters. All that
was required of the ltering in order to arrive at (10.9) from (10.8) was that it
remove frequencies above 4f0 . This mild requirement is accomplished even by
the integrator alone.
Exercise 10.7. The code in pllsd.m is simplied in the sense that the received
signal rp contains just the unmodulated carrier. Implement a more realistic
scenario by combining pulrecsig.m to include a binary message sequence,
pllpreprocess.m to create rp, and pllsd.m to recover the unknown phase oset
of the carrier.

202

Chapter 10. Carrier Recovery

rp(kTs)

cos(4f0kTs + 2[k])

LPF

[k]

sin(4f0kTs + 2[k])

2
Figure 10.6 A block diagram of the squared-dierence phase-tracking algorithm (10.7).
The input rp (kTs ) is a preprocessed version of the received signal as shown in
Figure 10.3. The integrator block has a lowpass character, and is equivalent to a
sum and delay as shown in Figure 7.5.

Exercise 10.8. Using the default values in pulrecsig.m and pllpreprocess.m

results in a of zero. Exercise 10.2 provided several situations in which =
0. Modify pllsd.m to allow for nonzero , and verify the code for the cases
suggested in Exercise 10.2.
Exercise 10.9. Investigate how the SD algorithm performs when the received
signal contains pulse-shaped 4-PAM data. Can you choose parameters so that
?
Exercise 10.10. Consider the sampled cosine wave
x(kTs ) = cos(2f0 kTs + ),
where the frequency f0 is known but the phase is not. Form
v(kTs ) = x(kTs ) cos(2f0 kTs + (k))
using the current (i.e., at the kth sample instant) estimate (k) of and dene
the candidate performance function
J() = LPF{v2 (kTs )},
where the cuto frequency of the ideal lowpass lter is 0.8f0 .
a. Does minimizing or maximizing J() result in = ? Justify your answer.
b. Develop a small-stepsize gradient-descent algorithm for updating (k) to estimate . Be certain that all of the signals needed to implement this algorithm
are measurable quantities. For example, x is directly measurable, but is not.
c. Determine the range of the initial guesses for (k) in the algorithm of part (b)
that will lead to the desired convergence to given a suitably small stepsize.

10.3

The Phase-Locked Loop

Perhaps the best-loved method of phase tracking is known as the phase-locked
loop (PLL). This section shows that the PLL can be derived as an adaptive

Carrier Recovery

203

element ascending the gradient of a simple performance function. The key idea
is to modulate the (processed) received signal rp (t) of Figure 10.3 down to DC,
using a cosine of known frequency 2f0 and phase 2 + . After ltering to remove
the high-frequency components, the magnitude of the DC term can be adjusted
by changing the phase. The value of that maximizes the DC component is the
same as the phase of rp (t).
To be specic, let
1
(10.10)
LPF{rp (kTs ) cos(4f0 kTs + 2 + )}.
2
Using the cosine product relationship (A.9) and the denition of rp (t) from (10.3)
under the assumption that fc = f0 ,
JP LL () =

1
LPF{cos(4f0 kTs + 2 + ) cos(4f0 kTs + 2 + )}
2
1
= LPF{cos(2 2) + cos(8f0 kTs + 2 + 2 + 2)}
4
1
1
= LPF{cos(2 2)} + LPF{cos(8f0 kTs + 2 + 2 + 2)}
4
4
1
(10.11)
cos(2 2),
4
assuming that the cuto frequency of the lowpass lter is well below 4f0 . This
is shown in the middle plot of Figure 10.5 and is the same as JSD (), except
for a constant and a sign. The sign change implies that, while JSD () needs to
be minimized to nd the correct answer, JP LL () needs to be maximized. The
substantive dierence between the SD and the PLL performance functions lies
in the way that the signals needed in the algorithm are extracted.
Assuming a small stepsize, the derivative of (10.10) with respect to at time
k can be approximated (using (G.12)) as

d LPF{rp (kTs ) cos(4f0 kTs + 2 + )} 

d
=[k]

#

drp (kTs ) cos(4f0 kTs + 2 + ) 
LPF

d
=[k]
JP LL () =

= LPF{rp (kTs ) sin(4f0 kTs + 2[k] + )}.

The corresponding adaptive element,
[k + 1] = [k] LPF{rp (kTs ) sin(4f0 kTs + 2[k] + )},

(10.12)

is shown in Figure 10.7. Observe that the sign of the derivative is preserved in
the update (rather than its negative), indicating that the algorithm is searching
for a maximum of the error surface rather than a minimum. One dierence
between the PLL and SD algorithms is clear from a comparison of Figures 10.6
and 10.7. The PLL requires one fewer oscillator (and one fewer addition block).
Since the performance functions JSD () and JP LL () are eectively the same,
the performance characteristics of the two are roughly equivalent.

204

Chapter 10. Carrier Recovery

rp(kTs)

e[k]
LPF F(z)

[k]

sin(4f0kTs + 2[k] + )
2
Figure 10.7 A block diagram of the digital phase-locked-loop algorithm (10.12). The

input signal rp (kTs ) has already been preprocessed to emphasize the carrier as in
Figure 10.3. The sinusoid mixes with the input and shifts the frequencies; after the
LPF only the components near DC remain. The loop adjusts to maximize this
low-frequency energy.

Suppose that fc is the frequency of the transmitter and f0 is the assumed

frequency at the receiver (with f0 close to fc ). The following program simulates
(10.12) for time seconds. Note that the firpm lter creates an h with a zero
phase at the center frequency and so is set to zero.
Listing 10.4. pllconverge.m simulate Phase-Locked Loop
Ts =1/10000; time =1; t=Ts : Ts : time ;
% time v e c t o r
f c =1000; p h o f f = 0.8;
% c a r r i e r f r e q and phase
rp=cos (4 * pi * f c * t +2* p h o f f ) ;
% s i m p l i f i e d rec d s i g n a l
f l =100; f f =[0 . 0 1 . 0 2 1 ] ; f a =[1 1 0 0 ] ;
h=f i r p m ( f l , f f , f a ) ;
% LPF d e s i g n
mu= . 0 0 3 ;
% algorithm s t e p s i z e
f 0 =1000;
% f r e q at r e c e i v e r
t h e t a=zeros ( 1 , length ( t ) ) ; t h e t a ( 1 ) = 0 ;
% i n i t i a l i z e estimates
z=zeros ( 1 , f l +1);
% i n i t i a l i z e LPF
f o r k =1: length ( t )1
% z contains past inputs
z =[ z ( 2 : f l +1) , rp ( k ) * s i n (4 * pi * f 0 * t ( k)+2 * t h e t a ( k ) ) ] ;
update=f l i p l r ( h ) * z ;
% new o u t p u t o f LPF
t h e t a ( k+1)= t h e t a ( k)mu* update ;
% a l g o r i t h m update
end

Figures 10.8(a) and (b) show the output of the program when f0 = fc and
f0 = fc , respectively. When the frequencies are the same, converges to a region
about the correct phase oset and wiggles about, with a size proportional
to the size of and dependent on details of the LPF. When the frequencies
are not the same, has a denite trend (the simulation in Figure 10.8(b) used
f0 = 1000 Hz and fc = 1001 Hz). Can you gure out how the slope of relates
to the frequency oset? The caption in Figure 10.8(b) provides a hint. Can you
imagine how the PLL might be used to estimate the frequency as well as to nd
the phase oset? These questions, and more, will be answered in Section 10.6.
Exercise 10.11. Use the preceding code to play with the phase-locked-loop
algorithm. How does aect the convergence rate? How does aect the oscillations in ? What happens if is too large (say = 1)? Does the convergence
speed depend on the value of the phase oset?

Phase offset

Carrier Recovery

0.2

0.4

205

0.6

0.8
0

0.2

0.4
0.6
Time
(a)

0.8

4
6
Time
(b)

10

Figure 10.8 Using the PLL, the estimates in (a) converge to a region about the
phase oset , and then oscillate. When the frequency estimate is incorrect as in (b),
becomes a line whose slope is proportional to the frequency dierence.

Exercise 10.12. In pllconverge.m, how much ltering is necessary? Reduce

the length of the lter. Does the algorithm still work with no LPF? Why? How
does your lter aect the convergent value of the algorithm? How does your lter
aect the tracking of the estimates when f0 = fc ?
Exercise 10.13. The code in pllconverge.m is simplied in the sense that the
received signal rp contains just the unmodulated carrier. Implement a more realistic scenario by combining pulrecsig.m to include a binary message sequence,
pllpreprocess.m to create rp, and pllconverge.m to recover the unknown
phase oset of the carrier.
Exercise 10.14. Using the default values in pulrecsig.m and pllpreprocess.m
results in a of zero. Exercise 10.2 provided several situations in which = 0.
Modify pllconverge.m to allow for nonzero , and verify the code on the cases
suggested in Exercise 10.2.
Exercise 10.15. TRUE or FALSE: The optimum settings of phase recovery for
a PLL operating on a preprocessed (i.e., squared and narrowly bandpass-ltered
at twice the carrier frequency) received PAM signal are unaected by the channel
transfer function outside a narrow band around the carrier frequency.
Exercise 10.16. Investigate how the PLL algorithm performs when the received
signal contains pulse-shaped 4-PAM data. Can you choose parameters so that
?
Exercise 10.17. Many variations on the basic PLL theme are possible. Letting u(kTs ) = rp (kTs ) cos(4f0 kTs + ), the preceding PLL corresponds to a performance function of JP LL () = LPF{u(kTs)}. Consider the alternative J() =

206

Chapter 10. Carrier Recovery

LPF{u2 (kTs )}, which leads directly to the algorithm5

#


du(kTs ) 
,
[k + 1] = [k] LPF u(kTs )
d =[k]
which is
[k + 1] = [k] LPF{rp2 (kTs ) sin(4f0 kTs + 2[k]) cos(4f0 kTs + 2[k])}.
a. Modify the code in pllconverge.m to play with this variation on the PLL.
Try a variety of initial values theta(1). Are the convergent values always the
same as with the PLL?
b. How does aect the convergence rate?
c. How does aect the oscillations in ?
d. What happens if is too large (say = 1)?
e. Does the convergence speed depend on the value of the phase oset?
f. What happens when the LPF is removed (set equal to unity)?
g. Draw the corresponding error surface.
Exercise 10.18. Consider the alternative performance function J() = |u(kTs )|,
where u(kTs ) is dened in Exercise 10.17. Derive the appropriate adaptive element, and implement it by imitating the code in pllconverge.m. In what ways
is this algorithm better than the standard PLL? In what ways is it worse?
The PLL can be used to identify the phase oset of the carrier. It can be
derived as a gradient descent on a particular performance function, and can
be investigated via simulation (with variants of pllconverge.m, for instance).
When the phase oset changes, the PLL can track the changes up to some
maximum rate. Conceptually, tracking a small frequency oset is identical to
tracking a changing phase, and Section 10.6 investigates how to use the PLL as
a building block for the estimation of frequency osets. Section 10.6.3 also shows
how a linearized analysis of the behavior of the PLL algorithm can be used
to concretely describe the convergence and tracking performance of the loop in
terms of the implementation of the LPF.

10.4

The Costas Loop

The PLL and the SD algorithms are two ways of synchronizing the phase at
the receiver to the phase at the transmitter. Both require that the received
signal be preprocessed (for instance, by a squaring nonlinearity and a BPF as
in Figure 10.3) in order to extract a clean version of the carrier, albeit at
twice the frequency and phase. An alternative approach operates directly on the
5

This is sensible because values of that minimize u2 (kTs ) also minimize u(kTs ).

207

Carrier Recovery

received signal r(kTs ) = s(kTs ) cos(2fc kTs + ) by reversing the order of the
processing: rst modulating to DC, then lowpass ltering, and nally squaring.
This reversal of operations leads to the performance function
JC () = avg{( LPF{r(kTs ) cos(2f0 kTs + )})2 }.

(10.13)

The resulting algorithm is called the Costas loop after its inventor J. P. Costas.
Because of the way in which the squaring nonlinearity enters JC (), it can operate
without preprocessing of the received signal as in Figure 10.3. To see why this
works, suppose that fc = f0 , and substitute r(kTs ) into (10.13):
JC () = avg{( LPF{s(kTs ) cos(2f0 kTs + ) cos(2f0 kTs + )})2 }.
Assuming that the cuto of the LPF is larger than the absolute bandwidth of
s(kTs ), and following the same logic as in (10.11) but with instead of 2, in
place of 2, and 2f0 kTs replacing 4f0 kTs shows that
LPF{s(kTs ) cos(2f0 kTs + ) cos(2f0 kTs + )}
1
(10.14)
= LPF{s(kTs )} cos( ).
2
Substituting (10.14) into (10.13) yields

2 #
1
1
s(kTs ) cos( )
JC () = avg
= avg{s2 (kTs ) cos2 ( ))}
2
4

1 2
cos2 ( ),
s
4 avg

where s2avg is the (xed) average value of the square of the data sequence s(kTs ).
Thus JC () is proportional to cos2 ( ). This performance function is plotted
(for an unknown phase oset of = 0.8) in the bottom part of Figure 10.5.
Like the error surface for the PLL (the middle plot), this achieves a maximum
when the estimate is equal to . Other maxima occur at + n for integer n.
In fact, except for a scaling and a constant, this is the same as JP LL because
cos2 ( ) = 12 (1 + cos(2 2)), as shown using (A.4).
The Costas loop can be implemented as a standard adaptive element (10.5).
The derivative of JC () is approximated by swapping the order of the dierentiation and the averaging (as in (G.12)), applying the chain rule, and then
swapping the derivative with the LPF. Thus dJC ()/d can be approximated as


d LPF{r(kTs ) cos(2f0 kTs + )}2
avg
d


d LPF{r(kTs ) cos(2f0 kTs + )}
= 2 avg LPF{r(kTs ) cos(2f0 kTs + )}
d


{dr(kTs ) cos(2f0 kTs + )}
2 avg LPF{r(kTs ) cos(2f0 kTs + )} LPF
d
= 2 avg{ LPF{r(kTs ) cos(2f0 kTs + )} LPF{r(kTs ) sin(2f0 kTs + )}}.

208

Chapter 10. Carrier Recovery

2cos(2f0kTs + [k])
Figure 10.9 The Costas loop is a

LPF

r(kTs)

[k]

LPF

phase-tracking algorithm based

on the performance function
(10.13). The input need not be
preprocessed (as is required by
the PLL).

2sin(2f0kTs + [k])

Accordingly, an implementable version of the Costas loop can be built as


dJC () 
[k + 1] = [k] +
d =[k]
= [k] avg{LPF{r(kTs ) cos(2f0 kTs + [k])}
LPF{r(kTs ) sin(2f0 kTs + [k])}}.
This is diagrammed in Figure 10.9, leaving o the rst (or outer) averaging
operation (as is often done), since it is redundant given the averaging eect of
the two LPFs and the averaging eect inherent in the small-stepsize update.
With this averaging removed, the algorithm is
[k + 1] = [k] LPF{r(kTs ) cos(2f0 kTs + [k])}
LPF{r(kTs ) sin(2f0 kTs + [k])}.

(10.15)

Basically, there are two paths. The upper path modulates by a cosine and then
lowpass lters to create (10.14), while the lower path modulates by a sine wave
and then lowpass lters to give s(kTs ) sin( ). These combine to give the
equation update, which is integrated to form the new estimate of the phase. The
latest phase estimate is then fed back (this is the loop in Costas loop) into
the oscillators, and the recursion proceeds.
Suppose that a 4-PAM transmitted signal r is created as in pulrecsig.m
(from page 194) with carrier frequency fc=1000. The Costas-loop phase-tracking
method (10.15) can be implemented in much the same way that the PLL is
implemented in pllconverge.m.
Listing 10.5. costasloop.m simulate the Costas loop with input from pulrecsig.m
r=r s c ;
f l =500; f f =[0 . 0 1 . 0 2 1 ] ; f a =[1 1 0 0 ] ;
h=f i r p m ( f l , f f , f a ) ;
mu= . 0 0 3 ;
f 0 =1000;
t h e t a=zeros ( 1 , length ( t ) ) ; t h e t a ( 1 ) = 0 ;
z s=zeros ( 1 , f l +1); zc=zeros ( 1 , f l +1);
f o r k =1: length ( t )1

% r s c from p u l r e c s i g .m
%
%
%
%
%
%

LPF d e s i g n
algorithm s t e p s i z e
f r e q at r e c e i v e r
estimate vector
b u f f e r s f o r LPFs
z contains past inputs

Carrier Recovery

209

3
2
1
Phase offset

Figure 10.10 Depending on

where it is initialized, the

estimates made by the
Costas-loop algorithm
converge to n. For this
plot, the unknown was
0.8, and there were 50
dierent initializations.

0
1
2
3
4
0

0.4

0.8

1.2

1.6

Time

z s =[ z s ( 2 : f l +1) , 2 * r ( k ) * s i n (2 * pi * f 0 * t ( k)+ t h e t a ( k ) ) ] ;
z c =[ z c ( 2 : f l +1) , 2 * r ( k ) * cos (2 * pi * f 0 * t ( k)+ t h e t a ( k ) ) ] ;
l p f s=f l i p l r ( h ) * zs ; l p f c=f l i p l r ( h ) * zc ; % o u t p u t o f f i l t e r s
t h e t a ( k+1)= t h e t a ( k)mu* l p f s * l p f c ;
% a l g o r i t h m update
end

Typical output of costasloop.m is shown in Figure 10.10, which shows the

evolution of the phase estimates for 50 dierent starting values theta(1). Some
of these converge to = 0.8, and others converge to nearby multiples. These
stationary points occur at all the maxima of the error surface (the bottom plot
in Figure 10.5).
When the frequency is not exactly known, the phase estimates of the Costas
algorithm try to follow. For example, in Figure 10.11, the frequency of the carrier
is fc = 1000, while the assumed frequency at the receiver was f0 = 1000.1. Fifty
dierent starting points are shown, and in all cases the estimates converge to a
line. Section 10.6 shows how this linear phase motion can be used to estimate
the frequency dierence.
Exercise 10.19. Use the preceding code to play with the Costas-loop algorithm.
a.
b.
c.
d.

How does the stepsize mu aect the convergence rate?

What happens if mu is too large (say mu=1)?
Does the convergence speed depend on the value of the phase oset?
When there is a small frequency oset, what is the relationship between the
slope of the phase estimate and the frequency dierence?

Exercise 10.20. How does the lter h inuence the performance of the Costas
loop?

210

Chapter 10. Carrier Recovery

6
4

Phase offset

2
Figure 10.11 When the
frequency of the carrier is
unknown at the receiver,
the phase estimates
converge to a line.

0
2
4

0.4

0.8

1.2

1.6

Time

a. Try fl=1000, 30, 10, 3.

b. Remove the LPFs completely from costasloop.m. How does this aect the
convergent values and the tracking performance?
Exercise 10.21. Oscillators that can adjust their phase in response to an input
signal are more expensive than free-running oscillators. Figure 10.12 shows an
alternative implementation of the Costas loop.
a. Show that this is actually carrying out the same calculations (albeit in a
dierent order) as the implementation in Figure 10.9.
b. Write a simulation (or modify costasloop.m) to implement this alternative.
Exercise 10.22. Reconsider the modied PLL of Exercise 10.17. This algorithm
also incorporates a squaring operation. Does it require the preprocessing step of
Figure 10.3? Why?

Exercise 10.23. TRUE or FALSE: Implementing a Costas-loop phase-recovery

scheme on the preprocessed version (i.e., squared and narrowly bandpass-ltered
at twice the carrier frequency) of a received PAM signal results in one and only
one local minimum in any 179 window of the adjusted phase.
In some applications, the Costas loop is considered a better solution than the
standard PLL because it can be more robust in the presence of noise.

10.5

Decision-Directed Phase Tracking

A method of phase tracking that works only in digital systems exploits the error
between the received value and the nearest symbol. For example, suppose that

211

Carrier Recovery

cos([k])

2cos(2f0kT)
LPF

r(kTs)

[k]

sin([k])
LPF

2sin(2f0kT)
cos([k])

Figure 10.12 An alternative implementation of the Costas loop trades o less

expensive oscillators for a more complex structure, as discussed in Exercise 10.21.

a 0.9 is received in a binary 1 system, suggesting that a +1 was transmitted.

Then the dierence between the 0.9 and the nearest symbol +1 provides information that can be used to adjust the phase estimate. This method is called
decision-directed (DD) because the decisions (the choice of the nearest allowable symbol) direct (or drive) the adaptation.
To see how this works, let s(t) be a pulse-shaped signal created from a message
in which the symbols are chosen from some (nite) alphabet. At the transmitter,
s(t) is modulated by a carrier at frequency fc with unknown phase , creating
the signal r(t) = s(t) cos(2fc t + ). At the receiver, this signal is demodulated
by a sinusoid and then lowpass ltered to create
x(t) = 2 LPF{s(t) cos(2fc t + ) cos(2f0 t + )}.

(10.16)

As shown in Chapter 5, when the frequencies (f0 and fc ) and phases ( and )
are equal, then x(t) = s(t). In particular, x(kTs ) = s(kTs ) at the sample instants
t = kTs , where the s(kTs ) are elements of the alphabet. On the other hand, if
= , then x(kTs ) will not be a member of the alphabet. The dierence between
what x(kTs ) is and what it should be can be used to form a performance function
and hence a phase-tracking algorithm. A quantization function Q(x) is used to
nd the nearest element of the symbol alphabet.
The performance function for the decision-directed method is
JDD () =

1
avg{(Q(x[k]) x[k])2 }.
4

(10.17)

212

Chapter 10. Carrier Recovery

2cos(2f0kTs + [k])
x(kT)

LPF
r(kTs)

Figure 10.13 The

Q( )

M
LPF

.
x(kT)

+
[k]

M
2sin(2f0kTs + [k])

decision-directed
phase-tracking
algorithm (10.18)
operates on the
dierence between the
soft decisions and the
hard decisions to drive
the updates of the
adaptive element.

This can be used as the basis of an adaptive element by using the approximation
(G.12) to calculate

#


dJDD ()
1
1
dx[k]
d (Q(x[k]) x[k])2
avg
= avg (Q(x[k]) x[k])
,
d
4
d
2
d
which assumes that the derivative of Q with respect to is zero. The derivative
of x[k] can similarly be approximated as (recall that x[k] = x(kTs ) = x(t)|t=kTs
is dened in (10.16))
dx[k]
2 LPF{r[k] sin(2fc kTs + )}.
d
Thus, the decision-directed algorithm for phase tracking is
[k + 1] = [k] avg{(Q(x[k]) x[k]) LPF{r[k] sin(2f0 kTs + [k])}}.
Suppressing the (redundant) outer averaging operation gives
[k + 1] = [k] (Q(x[k]) x[k]) LPF{r[k] sin(2f0 kTs + [k])},

(10.18)

which is shown in block diagram form in Figure 10.13.

Suppose that a 4-PAM transmitted signal rsc is created as in pulrecsig.m
(from page 194) with oversampling factor M=20 and carrier frequency fc=1000.
Then the DD phase-tracking method (10.18) can be simulated.
Listing 10.6. plldd.m decision-directed phase tracking
f l =100; f b e =[0 . 2 . 3 1 ] ;
damps=[1 1 0 0 ] ; h=f i r p m ( f l , f b e , damps ) ;
f z c=zeros ( 1 , f l +1); f z s=zeros ( 1 , f l +1);
t h e t a=zeros ( 1 ,N ) ; t h e t a (1 )= 0 . 9 ;
mu= . 0 3 ; j =1; f 0=f c ;
f o r k =1: length ( r s c )
c c =2* cos (2 * pi * f 0 * t ( k)+ t h e t a ( j ) ) ;
s s =2* s i n (2 * pi * f 0 * t ( k)+ t h e t a ( j ) ) ;
r c=r s c ( k ) * c c ; r s=r s c ( k ) * s s ;
f z c =[ f z c ( 2 : f l +1) , r c ] ;
f z s =[ f z s ( 2 : f l +1) , r s ] ;

%
%
%
%
%

LPF p a r a m e te r s
& impulse response
initialize filters
i n i t i a l phase
algorithm s t e p s i z e

%
%
%
%

c o s i n e f o r demod
s i n e f o r demod
do th e demods
s t a t e s f o r LPFs

Carrier Recovery

x ( k)= f l i p l r ( h ) * f z c ;
x d e r=f l i p l r ( h ) * f z s ;
i f mod ( 0 . 5 * f l +M/2k ,M)==0
qx=q u a n t a l p h ( x ( k ) , [ 3 , 1 , 1 , 3 ] ) ;
t h e t a ( j +1)= t h e t a ( j )mu * ( qxx ( k ) ) * x d e r ;
j=j +1;
end
end

%
%
%
%
%

213

LPF g i v e s x
& its derivative
downsample f o r t i m i n g
q u a n t i z e to a l p h a b e t
a l g o r i t h m update

The same lowpass lter is used after demodulation with the cosine (to create
x) and with the sine (to create its derivative xder). The ltering is done using
the time-domain method (the fourth method presented in waystofilt.m on page
143) because the demodulated signals are unavailable until the phase estimates
are made. One subtlety in the decision-directed phase-tracking algorithm is that
there are two time scales involved. The input, oscillators, and LPFs operate at
the faster sampling rate Ts , while the algorithm update (10.18) operates at the
slower symbol rate T . The correct relationship between these is maintained in
the code by the mod function, which picks one out of each M Ts -rate sampled data
points.
The output of plldd.m (for many dierent starting values) is shown in Figure
10.14. For initializations near the correct answer = 1.0, the estimates converge to 1.0. Of course, there is the (by now familiar) ambiguity. But there
are also other values for which the DD algorithm converges. What are these
values?
As with any adaptive element, it helps to draw the error surface in order
to understand its behavior. In this case, the error surface is JDD () plotted
as a function of the estimates . The following code approximates JDD () by
averaging over N=1000 symbols drawn from the 4-PAM alphabet.
Listing 10.7. plldderrsys.m error surface for decision-directed phase tracking
N=1000;
m=pam (N, 4 , 5 ) ;
p h i = 1.0;
theta = 2:.01:6;
f o r k =1: length ( t h e t a )
x=m* cos ( phit h e t a ( k ) ) ;
qx=q u a n t a l p h ( x , [ 3 , 1 , 1 , 3 ] ) ;
j t h e t a ( k )=( qx x ) * ( qx x ) /N;
end
plot ( th e ta , j t h e t a )

%
%
%
%
%
%
%
%

a v e r a g e o v e r N symbols ,
u s e 4PAM symbols
unknown phase o f f s e t p h i
g r i d f o r phase e s t i m a t e s
f o r each p o s s i b l e t h e t a
f i n d x with t h i s t h e t a
q( x) f o r t h i s theta
J f o r t h i s theta

The output of plldderrsys.m is shown in Figure 10.15. First, the error surface
inherits a basic periodicity6 in of period 2 from the cosine function. Within
each 2 repetition, there are six minima, two of which are broad and deep. One of
these corresponds to the correct phase at = 1.0 2n and the other (at =
1.0 + 2n) corresponds to the situation in which the cosine takes on a value
6

Actually, the error surface is periodic with period , but the behavior of the algorithm is
easier to understand by considering the fundamental periods in pairs.

Chapter 10. Carrier Recovery

Phase estimates

4
3
Figure 10.14 The decision-directed

tracking algorithm is adapted to

locate a phase oset of = 1.0.
Many of the dierent initializations
converge to = 1.0 n, but
there are also other convergent
values.

2
1
0

1
2

Time

1
0.8
Cost JDD()

214

Figure 10.15 The error surface for

the DD phase-tracking algorithm
(10.18) has several minima within
each 2 repetition. The phase
estimates will typically converge to
the closest of these minima.

0.6
0.4
0.2
0
2

1
2
3
Phase estimates

of 1. This inverts each data symbol: 1 is mapped to 1, and 3 is mapped to

3. The other four occur near multiples of 3/8 1.0 and 5/8 1.0, which
correspond to values of the cosine that jumble the data sequence in various ways.
The implication of this error surface is clear: there are many values to which
the decision-directed method may converge. Only some of these correspond to
desirable answers. Thus, the DD method is local in the same way that the steepest
descent minimization of the function (6.8) (in Section 6.6) depends on the initial
value of the estimate. If it is possible to start near the desired answer, then
convergence can be assured. However, if no good initialization is possible, the
method may converge to one of the undesirable minima. This suggests that
the decision-directed method can perform acceptably in a tracking mode (when
following a slowly varying phase), but would perform worse than the alternatives
at start-up, when nothing is known about the correct value of the phase.

Carrier Recovery

215

Exercise 10.24. Use the code in plldd.m to play with the DD algorithm.
a. How large can the stepsize be made?
b. Is the LPF of the derivative really needed?
c. How crucial is it to the algorithm to pick the correct timing? Examine this
question by choosing incorrect j at which to evaluate x.
d. What happens when the assumed frequency f0 is not the same as the frequency fc of the carrier?
Exercise 10.25. The direct calculation of dx(kTs )/d as a ltered version of
(10.16) is only one way to calculate the derivative. Replace this using a numerical
approximation (such as the forward or backward Euler, or the trapezoidal rule).
Compare the performance of your algorithm with that of plldd.m.
Exercise 10.26. Consider the DD phase-tracking algorithm when the message
alphabet is binary 1.
a. Modify plldd.m to simulate this case.
b. Modify plldderrsys.m to draw the error surface. Is the DD algorithm better
(or worse) suited to the binary case than the 4-PAM case?
Exercise 10.27. Consider the DD phase-tracking algorithm when the message
alphabet is 6-PAM.
a. Modify plldd.m to simulate this case.
b. Modify plldderrsys.m to draw the error surface. Is the DD algorithm better
(or worse) suited to 6-PAM than to 4-PAM?
Exercise 10.28. What happens when the number of inputs used to calculate
the error surface is too small? Try N = 100, 10, 1. Can N be too large?
Exercise 10.29. Investigate how the error surface depends on the input signal.
a. Draw the error surface for the DD phase-tracking algorithm when the inputs
are binary 1.
b. Draw the error surface for the DD phase-tracking algorithm when the inputs
are drawn from the 4-PAM constellation, for the case in which the symbol 3
never occurs.
Exercise 10.30. TRUE or FALSE: Decision-directed phase recovery can exhibit
local minima of dierent depths.
Exercise 10.31. This problem uses the B 3IG to test the carrier-recovery algorithm of your choice.
a. Currently, the receiver portion of the script in BigIdeal.m knows the phase
of the carrier. Add a carrier-recovery algorithm of your choice to BigIdeal.m

216

Chapter 10. Carrier Recovery

so that the receiver can accommodate unknown carrier phases. Run the modied code and use the carrier-recovery algorithm to estimate the phase oset
. State the carrier-recovery method used and plot the tracking of . If the
receiver employs preprocessing, take care to design the BPF appropriately.
b. Modify the B 3IG so that a phase oset is introduced. In BigTransmitter.m,
set the variable phi=-0.5, and add a phase oset
x_rf=x.*cos(2*pi*rfParams.f_if_tx*...
[1:length(x)]*rfParams.T_t_tx/M+p_noise+phi);
where upconversion is performed. Rerun the code. How does the behavior of
your algorithm dier? Include plots to justify your claim.
c. Repeat part (b) for the case when phi=1.2.

10.6

Frequency Tracking
The problems inherent in even a tiny dierence in the frequency of the carrier
at the transmitter and the assumed frequency at the receiver are shown in (5.5)
and illustrated graphically in Figure 9.19 on page 185. Since no two independent
oscillators are ever exactly aligned, it is important to nd ways of estimating the
frequency from the received signal. The direct method of Section 10.6.1 derives
an algorithm based on a performance function that uses a square dierence in
the time domain. Unfortunately, this does not work well, and its failure can be
traced to the shape of the error surface.
Section 10.6.2 begins with the observation (familiar from Figures 10.8(b) and
10.11) that the estimates of phase made by the phase-tracking algorithms over
time lie on a line whose slope is proportional to the dierence in frequency
between the modulating and the demodulating oscillators. This slope contains
valuable information that can be exploited to indirectly estimate the frequency.
The resulting dual-loop element is actually a special case of a more general (and
ultimately simpler) technique that puts an integrator in the forward part of the
loop in the PLL. This is detailed in Section 10.6.3.

10.6.1

Direct Frequency Estimation

Perhaps the simplest setting in which to begin frequency estimation is to assume
that the received signal is r(t) = cos(2fc t), where fc is unknown. By analogy
with the squared-dierence method of phase estimation in Section 10.2, a reasonable strategy is to try to choose f0 so as to minimize
J(f0 ) =

1
LPF{(r(t) cos(2f0 t))2 }.
2

(10.19)

Carrier Recovery

217

Frequency estimates f0

102

Figure 10.16 The

frequency-estimation
algorithm (10.20) appears to
function well at rst. But,
over time, the estimates
diverge from the desired
answer.

100

98

Time

12

16

Following a gradient strategy for updating the estimates f0 leads to the algorithm

dJ(f0 ) 
f0 [k + 1] = f0 [k]
(10.20)
df0 f0 =f0 [k]
= f0 [k] LPF{2kTs (r(kTs ) cos(2kTs f0 [k])) sin(2kTs f0 [k])}.
How well does this algorithm work? First, observe that the update is multiplied by 2kTs. (This arises from application of the chain rule when taking the
derivative of sin(2kTs f0 [k]) with respect to f0 [k].) This factor increases continuously, and acts like a stepsize that grows over time. Perhaps the easiest way to
make any adaptive element fail is to use a stepsize that is too large; the form of
this update ensures that eventually the stepsize will be too large.
Putting on our best engineering hat, let us just remove this oending term,
and go ahead and simulate the method.7 At rst glance it might seem that the
method works well. Figure 10.16 shows 20 dierent starting values. All 20 appear
to converge nicely within one second to the unknown frequency value at fc=100.
But then something strange happens: one by one, the estimates diverge. In the
gure, one peels o at about 6 seconds, and one at about 17 seconds. Simulations
can never prove conclusively that an algorithm is good for a given task, but if
even simplied and idealized simulations function poorly, it is a safe bet that the
algorithm is somehow awed. What is the aw in this case?
Recall that error surfaces are often a good way of picturing the behavior
of gradient-descent algorithms. Expanding the square and using the standard
identities (A.4) and (A.9), J(f0 ) can be rewritten

1
1
J(f0 ) = LPF 1 + cos(4fc t) cos(2(fc f0 )t)
2
2

1
+ cos(4f0 t) cos(2(fc + f0 )t)
2
1 1
(10.21)
= LPF{cos(2(fc f0 )t)},
2 2
7

The code is available in the program pllfreqest.m on the website.

218

Chapter 10. Carrier Recovery

0.5
J( f0 )

Figure 10.17 The error surface

fc
Frequency estimate f0

corresponding to the frequency

estimation performance function
(10.19) is at everywhere except
for a deep crevice at the correct
answer f0 = fc .

assuming that the cuto frequency of the lowpass lter is less than fc and that
f0 fc . At the point where f0 = fc , J(f0 ) = 0. For any other value of f0 , however, as time progresses, the cosine term undulates up and down with an average
value of zero. Hence J(f0 ) averages to 12 for any f0 = fc ! This pathological situation is shown in Figure 10.17.
When f0 is far from fc , this analysis does not hold because the LPF no longer
removes the rst two cosine terms in (10.21). Somewhat paradoxically, the algorithm behaves well until the answer is nearly correct. Once f0 fc , the error
surface attens, and the estimates wander around. There is a slight possibility that it might accidently fall into the exact correct answer, but simulations
suggest that such luck is rare. Oh well, never mind.

10.6.2

Indirect Frequency Estimation

Because the direct method of the previous section is unreliable, this section
pursues an alternative strategy based on the observation that the phase estimates
of the PLL converge to a line that has a slope proportional to the dierence
between the actual frequency of the carrier and the frequency that is assumed at
the receiver. (Recall Figures 10.8(b) and 10.11.) The indirect method cascades
two PLLs: the rst nds this line (and hence indirectly species the frequency),
the second converges to a constant appropriate for the phase oset.
The scheme is pictured in Figure 10.18. Suppose that the received signal has
been preprocessed to form rp (t) = cos(4fc t + 2). This is applied to the inputs
of two PLLs.8 The top PLL functions exactly as expected from previous sections:
if the frequency of its oscillator is 2f0 , then the phase estimates 21 converge to
a ramp with slope 2(f0 fc ), that is,
1 (t) 2(fc f0 )t + b,
8

Or two SD phase-tracking algorithms or two Costas loops, though in the latter case the
squaring preprocessing is unnecessary.

Carrier Recovery

rp(kTs)

e1[k]

LPF F1(z)

219

1[k]

sin(4f0kTs +21[k])
2
X

LPF F2(z)

e2[k]

2[k]

sin(4f0kTs +21[k] + 22[k])

2

1[k] + 2[k]

Figure 10.18 A pair of PLLs can eciently estimate the frequency oset at the

receiver. The parameter 1 in the top loop converges to a slope that corrects the
frequency oset and the parameter 2 in the bottom loop corrects the residual phase
oset. The sum 1 + 2 is used to drive the sinusoid in the carrier-recovery scheme.

where b is the y-intercept of the ramp. The 1 values are then added to 2 , the
phase estimate in the lower PLL. The output of the bottom oscillator is
sin(4f0 t + 21 (t) + 22 (t)) = sin(4f0 t + 4(fc f0 )t + 2b + 22 (t))
sin(4fc t + 2b + 22 (t)).
Eectively, the top loop has synthesized a signal that has the correct frequency
for the bottom loop. Accordingly, 2 (t) b. Since a sinusoid with frequency
2f0 t and phase 1 (t) + 2 (t) is indistinguishable from a sinusoid with frequency 2fc t and phase 2 (t), these values can be used to generate a sinusoid
that is aligned with rp (t) in both frequency and phase. This signal can then be
used to demodulate the received signal.
Some Matlab code to implement this dual PLL scheme is provided by
dualplls.m.
Listing 10.8. dualplls.m estimation of carrier via dual-loop structure
Ts =1/10000; time =5; t =0: Ts : time Ts ; % time v e c t o r
f c =1000; p h o f f =2;
% c a r r i e r f r e q and phase
rp=cos (4 * pi * f c * t +2* p h o f f ) ;
% p r e p r o c e s s e d s i g n a l rBPF
mu1 = . 0 1 ; mu2 = . 0 0 3 ;
% algorithm s t e p s i z e s
f 0 =1001;
% assumed f r e q a t r e c e i v e r
l e n t=length ( t ) ; th1=zeros ( 1 , l e n t ) ; % i n i t i a l i z e e s t i m a t e s
th2=zeros ( 1 , l e n t ) ; c a r e s t =zeros ( 1 , l e n t ) ;
f o r k =1: l e n t 1
% combine top PLL th1
th1 ( k+1)= th1 ( k)mu1 * rp ( k ) * s i n (4 * pi * f 0 * t ( k)+2 * th1 ( k ) ) ;
th2 ( k+1)= th2 ( k)mu2 * rp ( k ) * s i n (4 * pi * f 0 * t ( k)+2 * th1 ( k)+2 * th2 ( k ) ) ;
% with bottom PLL th2 to form e s t i m a t e o f p r e p r o c e s s e d s i g n a l
c a r e s t ( k)=cos (4 * pi * f 0 * t ( k )+2 * th1 ( k)+2 * th2 ( k ) ) ;
end

The output of this program is shown in Figure 10.19. The upper graph shows
that 1 , the phase estimate of the top PLL, converges to a ramp. The middle plot

Chapter 10. Carrier Recovery

20
0

20
Figure 10.19 The output of

Matlab program dualplls.m

shows the output of the rst PLL
converging to a line, which allows
the second PLL to converge to a
constant. The bottom gure shows
that this estimator can be used to
construct a sinusoid that is very
close to the (preprocessed) carrier.

0.3
0.2
0.1
0
0.1

error

220

1
0
1
0

Time

shows that 2 , the phase estimate of the bottom PLL, converges to a constant.
Thus the procedure is working. The bottom graph shows the error between the
preprocessed signal rp and a synthesized carrier carest. The parameters f0,
th1, and th2 can then be used to synthesize a cosine wave that has the correct
frequency and phase to demodulate the received signal.
It is clear from the top plot of Figure 10.19 that 1 converges to a line. What
line does it converge to? Looking carefully at the data generated by dualplls.m,
the line can be calculated explicitly. The two points at (2, 11.36) and (4, 23.93)
t a line with slope m = 6.28 and an intercept b = 1.21. Thus,
2(fc f0 ) = 6.28,
or fc f0 = 1. Indeed, this was the value used in the simulation. Reading the
nal converged value of 2 from the simulation shown in the middle plot gives
0.0627. b 0.0627 is 1.147, which is almost exactly away from 2, the value
used in phoff.
Exercise 10.32. Use the preceding code to play with the frequency estimator.
a. How far can f0 be from fc before the estimates deteriorate?
b. What is the eect of the two stepsizes mu? Should one be larger than other?
If so, which one?
c. How does the method fare when the input is noisy?
d. What happens when the input is modulated by pulse-shaped data instead of
by a simple sinusoid?
Exercise 10.33. Build a frequency estimator using two SD phase-tracking algorithms, rather than two PLLs. How does the performance change? Which do you
think is preferable?

Carrier Recovery

221

Exercise 10.34. Build a frequency estimator that incorporates the preprocessing of the received signal from Figure 10.3 (as coded in pllpreprocess.m).
Exercise 10.35. Build a frequency estimator using two Costas loops, rather than
two PLLs. How does the performance change? Which do you think is preferable?
Exercise 10.36. Investigate (via simulation) how the PLL functions when there
is white noise (using randn) added to the received signal. Do the phase estimates
become worse as the noise increases? Make a plot of the standard deviation of
the noise versus the average value of the phase estimates (after convergence).
Make a plot of the standard deviation of the noise versus the jitter in the phase
estimates.
Exercise 10.37. Repeat Exercise 10.36 for the dual SD algorithm.
Exercise 10.38. Repeat Exercise 10.36 for the dual Costas-loop algorithm.
Exercise 10.39. Repeat Exercise 10.36 for the dual DD algorithm.
Exercise 10.40. Investigate (via simulation) how the PLL functions when
there is intersymbol interference caused by a nonunity channel. Pick a channel
(for instance chan=[1, .5, .3, .1];) and incorporate this into the simulation
of the received signal. Using this received signal, are the phase estimates worse
when the channel is present? Are they biased? Are they more noisy?
Exercise 10.41. Repeat Exercise 10.40 for the dual Costas loop.
Exercise 10.42. Repeat Exercise 10.40 for the Costas-loop algorithm.
Exercise 10.43. Repeat Exercise 10.40 for the DD algorithm.

10.6.3

Generalized PLL
Section 10.6.2 showed that two loops can be used in concert to accomplish the
carrier-recovery task: the rst loop estimates frequency oset and the second
loop estimates the phase oset. This section shows an alternative structure that
folds one of the loops into the lowpass lter in the form of an IIR lter and
accomplishes both estimations at once. This is shown in Figure 10.20, which
looks the same as Figure 10.7 but with the FIR lowpass lter F (z) replaced by
an IIR lowpass lter G(z) = GN (z)/GD (z).
To see the operation of the generalized PLL concretely, pllgeneral.m implements the system of Figure 10.20 with GN (z) = 2 2z and GD (z) = z
1. The implementation of the IIR lter follows the time-domain method in
waystofiltIIR.m on page 145. When the frequency oset is zero (when f0=fc),

222

Chapter 10. Carrier Recovery

rp(kTs)

G(z)

[k+1]

z1

[k]

sin(4f0kTs + 2[k])
2
Figure 10.20 The general PLL can track both phase and frequency osets using an IIR

LPF G(z). This gure is identical to Figure 10.7, except that the FIR lowpass lter is
replaced by a suitable IIR lowpass lter. Linearizing this structure gives Figure 10.22.
Appendix F presents conditions on G(z) under which the tracking occurs.

theta converges to phoff. When there is a frequency oset (as in the default
values with f0=fc+0.1), theta converges to a line. Essentially, this value converges to 1 + 2 in the dual-loop structure of Figure 10.19.
Listing 10.9. pllgeneral.m the PLL with an IIR LPF
Ts =1/10000; time =1; t=Ts : Ts : time ;
f c =1000; p h o f f = 0 . 8 ;
rp=cos (4 * pi * f c * t +2* p h o f f ) ;
mu= . 0 0 3 ;
a =[1 1 ] ; l e n a=length ( a ) 1;
b=[2 2mu ] ; l e n b=length ( b ) ;
xvec=zeros ( l e n a , 1 ) ;
e v e c=zeros ( l e n b , 1 ) ;
f0 =1000.0;
t h e t a=zeros ( 1 , length ( t ) ) ; t h e t a ( 1 ) = 0 ;
f o r k =1: length ( t )1
e=rp ( k ) * s i n (4 * pi * f 0 * t ( k)+2 * t h e t a ( k ) ) ;
e v e c =[ e ; e v e c ( 1 : l en b 1 ) ] ;
x=a ( 2 : l e n a +1) * xvec+b * e v e c ;
xvec =[x ; xvec ( 1 : l e n a 1 , 1 ) ] ;
t h e t a ( k+1)= t h e t a ( k)+mu* x ;
end

%
%
%
%
%
%
%

time v e c t o r
c a r r i e r f r e q and phase
s i m p l i f i e d r ec d s i g n a l
algorithm s t e p s i z e
a u t o r e g r e s s i v e terms
moving a v e r a g e terms
initialize filters

% f r e q at r e c e i v e r
% i n i t i a l i z e estimates
%
%
%
%
%

i n p u t to f i l t e r
past values of inputs
output o f f i l t e r
past values of outputs
a l g o r i t h m update

All of the PLL structures are nonlinear, which makes them hard to analyze
exactly. One way to study the behavior of a nonlinear system is to replace nonlinear components with nearby linear elements. A linearization is valid only for
a small range of values. For example, the nonlinear function sin(x) is approximately x for small values. This can be seen by looking at sin(x) near the origin:
it is approximately a line with slope unity. As x increases or decreases away from
the origin, the approximation worsens.
The nonlinear components in a loop structure such as Figure 10.7 are the
oscillators and the mixers. Consider a nonideal lowpass lter F (z) with impulse
response f [k] that has a cuto frequency below 2f0 . The output of the LPF is
e[k] = f [k] sin(2[k] 2[k]), where represents convolution. With

and , sin(2[k])
2[k].
Thus the linearization is

f [k] sin(2[k])
f [k] 2[k].

(10.22)

Carrier Recovery

223

In words, applying an LPF to sin(x) is approximately the same as applying the

LPF to x, at least for suciently small x.
There are two ways to understand why the general PLL structure in Figure
10.20 works, both of which involve linearization arguments. The rst method linearizes Figure 10.20 and applies the nal-value theorem for Z-transforms. Details
can be found in Section F.4. The second method is to show that a linearization
of the generalized single-loop structure of Figure 10.20 is the same as a linearization of the dual-loop method of Figure 10.19, at least for certain values of the
parameters. Exercise 10.50 shows that, for certain special cases, the two strategies (the dual-loop and the generalized single-loop) behave similarly in the sense
that both have the same linearization.
Exercise 10.44. What is the largest frequency oset that pllgeneral.m can
track? How does theta behave when the tracking fails?
Exercise 10.45. Examine the eect of dierent lters in pllgeneral.m.
a. Show that the lter used in pllgeneral.m has a lowpass character.
b. Let a=[1 0]. What kind of IIR lter does this represent? How does the loop
behave for a frequency oset?
c. Can you choose b so that the loop fails?
d. Design an IIR a and b pair using cheby1 or cheby2. Test whether the loop
functions to track small frequency osets.
Exercise 10.46. Design the equivalent squared-dierence loop, combining the
IIR structure of pllgeneral.m with the squared-dierence objective function as
in pllsd.m.
Exercise 10.47. Design the equivalent general Costas-loop structure. Hint:
replace the FIR LPFs in Figure 10.9 with suitable IIR lters. Create a simulation to verify the operation of the method.
Exercise 10.48. The code in pllgeneral.m is simplied in the sense that the
received signal rp contains just the unmodulated carrier. Implement a more realistic scenario by combining pulrecsig.m to include a binary message sequence,
pllpreprocess.m to create rp, and pllgeneral.m to recover the unknown phase
oset of the carrier. Demonstrate that the same system can track a frequency
oset.
Exercise 10.49. Investigate how the method performs when the received signal
contains pulse-shaped 4-PAM data. Verify that it can track both phase and
frequency osets.

224

Chapter 10. Carrier Recovery

2[k]

F1(z)

e1[k]

1[k + 1]

1[k]

z1

2
+

F2(z)

e2[k]

2[k +1]

2[k]

z1

21[k] + 22[k]
2
Figure 10.21 A linearization of the dual-loop frequency tracker of Figure 10.18. For

small osets, this linear structure closely represents the behavior of the system.

2[k]

e[k]

GN(z)
GD(z)

[k + 1]

z1

[k]
Figure 10.22 This is the

linearization of Figure
10.20.
2

A
+

B
C

Figure 10.23 The left-hand block diagram with overlapping loops can be redrawn as

the right-hand split diagram.

Exercise 10.50. This exercise outlines the steps needed to show the conditions
under which the general PLL of Figure 10.20 and the dual-loop structure of
Figure 10.18 are the same (in the sense of having the same linearization).
a.
b.
c.
d.

Show that Figure 10.21 is the linearization of Figure 10.18.

Show that Figure 10.22 is the linearization of the general PLL of Figure 10.20.
Show that the two block diagrams in Figure 10.23 are the same.
Suppose that the lowpass lters are ideal: F1 (z) = F2 (z) = 1 for all frequencies
below the cuto and F1 (z) = F2 (z) = 0 at all frequencies above. Make the
following assignments:

, B=
, and C = 2.
A=
z 1 + 2
z1

Carrier Recovery

225

Using the block-diagram manipulation of part (c), show that Figures 10.21
and 10.22 are the same when GN (z) = (2z 2 + 2) and GD (z) = z 1.

For Further Reading

r J. P. Costas, Synchronous Communications, Proceedings of the IRE, vol.

44, no. 12, pp. 17131718, Dec. 1956.

r L. E. Franks, Carrier and Bit Synchronization in Data Communication

A Tutorial Review, IEEE Transactions on Communications, vol. 28, no. 8,
pp. 11071120, Aug. 1980.

11 Pulse Shaping and Receive Filtering

When the message is digital, it must be converted into an analog signal in order
to be transmitted. This conversion is done by the transmit or pulse-shaping
lter, which changes each symbol in the digital message into a suitable analog
pulse. After transmission, the receive lter assists in recapturing the digital
values from the received pulses. This chapter focuses on the design and specication of these lters.
The symbols in the digital input sequence w(kT ) are chosen from a nite set
of values. For instance, they might be binary 1, or they may take values from
a larger set such as the four-level alphabet 1, 3. As suggested in Figure 11.1,
the sequence w(kT ) is indexed by the integer k, and the data rate is one symbol
every T seconds. Similarly, the output m(kT ) assumes values from the same
alphabet as w(kT ) and at the same rate. Thus the message is fully specied at
times kT for all integers k. But what happens between these times, between kT
and (k + 1)T ? The analog modulation of Chapter 5 operates continuously, and
some values must be used to ll in the digital input between the samples. This
is the job of the pulse-shaping lter: to turn a discrete-time sequence into an
analog signal.
Each symbol w(kT ) of the message initiates an analog pulse that is scaled by
the value of the signal. The pulse progresses through the communications system,
and, if all goes well, the output (after the decision) should be the same as the
input, although perhaps with some delay. If the analog pulse is wider than the
time between adjacent symbols, the outputs from adjacent symbols may overlap,
a problem called intersymbol interference, which is abbreviated ISI. A series of
examples in Section 11.2 shows how this happens, and the eye diagram is used
in Section 11.3 to help visualize the impact of ISI.
What kinds of pulses minimize the ISI? One possibility is to choose a shape
that is one at time kT and zero at mT for all m = k. Then the analog waveform at time kT contains only the value from the desired input symbol, and no
interference from other nearby input symbols. These are called Nyquist pulses in

Pulse Shaping and Receive Filtering

p(t)
P( f )
Pulse
shaping

x(t)

hc(t)
Hc( f )
Channel

Interferers

Noise
n(t)

g(t)

Message
w(kT) {3, 1, 1, 3}

hR(t)
HR( f )
Receive
filter

227

y(t)
Decision
Reconstructed
message
m(kT) {3, 1, 1, 3}

Figure 11.1 System diagram of a baseband communication system emphasizing the

pulse shaping at the transmitter and the corresponding receive ltering at the
receiver.

Section 11.4. Yes, this is the same fellow who brought us the Nyquist sampling
theorem and the Nyquist frequency.
Besides choosing the pulse shape, it is also necessary to choose a receive lter
that helps decode the pulses. The received signal can be thought of as containing
two parts: one part is due to the transmitted signal and the other part is due to
the noise. The ratio of the powers of these two parts is a kind of signal-to-noise
ratio that can be maximized by choice of the pulse shape. This is discussed in
Section 11.5. The chapter concludes in Section 11.6 by considering pulse shaping
and receive lters that both provide a Nyquist pulse and maximize the signalto-noise ratio.
The transmit and receive lter designs rely on the assumption that all other
parts of the system are working well. For instance, the modulation and demodulation blocks have been removed from Figure 11.1, and the assumption is that
they are perfect: the receiver knows the correct frequency and phase of the carrier. Similarly, the downsampling block has been removed, and the assumption
is that this is implemented so that the decision device is a fully synchronized
sampler and quantizer. Chapter 12 examines methods of satisfying these synchronization needs, but, for now, they are assumed to be met. In addition, the
channel is assumed benign.

11.1

Spectrum of the Pulse: Spectrum of the Signal

Probably the major reason why the design of the pulse shape is important is
because the shape of the spectrum of the pulse dictates the spectrum of the
whole transmission. To see this, suppose that the discrete-time message sequence
w(kT ) is turned into the analog pulse train


w(kT ) t = kT
w(kT )(t kT ) =
(11.1)
wa (t) =
0
t = kT
k

as it enters the pulse-shaping lter. The response of the lter, with impulse
response p(t), is the convolution
x(t) = wa (t) p(t),

Pulse shape

Chapter 11. Pulse Shaping and Receive Filtering

Spectrum of the pulse shape

228

1
0.8
0.6
0.4
0.2
0
0

0.1

0.1

0.2

0.3 0.4 0.5 0.6

(a) Sample periods

0.7

0.8

0.9

0.9

102

Figure 11.2 The Hamming

pulse shape and its
magnitude spectrum.

100
102
104

0.2

0.3 0.4 0.5 0.6 0.7

(b) Normalized frequency

0.8

as suggested by Figure 11.1. Since the Fourier transform of a convolution is the

product of the Fourier transforms (from (A.40)), it follows that
X(f ) = Wa (f )P (f ).
Though Wa (f ) is unknown, this shows that X(f ) can have no energy at frequencies at which P (f ) vanishes. Whatever the spectrum of the message, the
transmission is directly scaled by P (f ). In particular, the support of the spectrum X(f ) is no larger than the support of the spectrum P (f ).
As a concrete example, consider the pulse shape used in Chapter 9, which
is the blip function shown in the top plot of Figure 11.2. The spectrum of
this pulse shape can readily be calculated using freqz, and this is shown at the
bottom plot of Figure 11.2. It is a kind of mild lowpass lter. The following code
generates a sequence of N 4-PAM symbols, and then carries out the pulse shaping
using the filter command.
Listing 11.1. pulsespec.m spectrum of a pulse shape
N=1000; w=pam (N, 4 , 5 ) ;
M=10; wup=zeros ( 1 ,N*M) ;
wup ( 1 :M:N*M)=w ;
ps=hamming (M) ;
x=f i l t e r ( ps , 1 , wup ) ;

%
%
%
%
%

4 l e v e l s i g n a l o f l e n g t h N
oversampling f a c t o r
o v e r s a m p l e by M
b l i p p u l s e o f width M
c o n v o l v e p u l s e shape with data

The program pulsespec.m represents the continuous-time or analog signal by oversampling both the data sequence and the pulse shape by a factor of M. This technique was discussed in Section 6.3, where an analog sine
wave sine100hzsamp.m was represented digitally at two sampling intervals, a
slow symbol interval T = M Ts and a faster rate (shorter interval) Ts representing the underlying analog signal. The pulse shape ps is a blip created by the
hamming function, and this is also oversampled at the same rate. The convolution of the oversampled pulse shape and the oversampled data sequence is
accomplished by the filter command. Typical output is shown in the top plot

Output of pulse
shaping filter

Pulse Shaping and Receive Filtering

3
2
1
0
1
2
3

Spectrum of the output

229

10
15
Symbols

20

0.2

0.3 0.4 0.5 0.6 0.7

Normalized frequency

0.8

104

25

102
100
102
104
0

0.1

0.9

Figure 11.3 The top plot shows a segment of the output x of the pulse-shaping lter.
The bottom plot shows the magnitude spectrum of x, which has the same general
contour as the spectrum of a single copy of the pulse. Compare this with the bottom
plot of Figure 11.2.

of Figure 11.3, which shows the analog signal over a time interval of about 25
symbols. Observe that the individual pulse shapes are clearly visible, one scaled
blip for each symbol.
The spectrum of the output x is plotted at the bottom of Figure 11.3. As
expected from the previous discussion, the spectrum X(f ) has the same contour
as the spectrum of the individual pulse shape in Figure 11.2.

11.2

Intersymbol Interference
There are two situations in which adjacent symbols may interfere with each other:
when the pulse shape is wider than a single symbol interval T , and when there is
a nonunity channel that smears nearby pulses, causing them to overlap. Both
of these situations are called intersymbol interference (ISI). Only the rst kind of
ISI will be considered in this chapter; the second kind is postponed until Chapter
13. Before tackling the general setup, this section provides an instructive example
where the ISI is caused by an overly wide pulse shape.

Example 11.1. Suppose that the pulse shape in pulsespec.m is stretched so

that its width is 3T . This triple-width Hamming pulse shape is shown in Figure
11.4, along with its spectrum. Observe that the spectrum has (roughly) one-third
the null-to-null bandwidth of the single-symbol-wide Hamming pulse. Since the

The pulse shape

Chapter 11. Pulse Shaping and Receive Filtering

1
0.8
0.6
0.4
0.2
0
0

Spectrum of the pulse shape

230

0.5

1
1.5
2
(a) Sample periods

2.5

102

Figure 11.4 The triple-width

Hamming pulse shape and its
magnitude spectrum, which is
drawn using freqz.

100
102
104
0

0.1

0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

(b) Normalized frequency

width of the spectrum of the transmitted signal is dictated by the width of the
spectrum of the pulse, this pulse shape is three times as parsimonious in its use
of bandwidth. More FDM users can be active at the same time.
As might be expected, this boon has a price. Figure 11.5 shows the output of
the pulse-shaping lter over a time of about 25 symbols. There is no longer a
clear separation of the pulse corresponding to one data point from the pulses of
its neighbors. The transmission is correspondingly harder to properly decode. If
the ISI caused by the overly wide pulse shape is too severe, symbol errors may
occur.
Thus, there is a trade-o. Wider pulse shapes can occupy less bandwidth,
which is always a good thing. On the other hand, a pulse shape like the Hamming
blip does not need to be very many times wider before it becomes impossible to
decipher the data because the ISI has become too severe. How much wider can it
be without causing symbol errors? The next section provides a way of picturing
ISI that answers this question. Subsequent sections discuss the practical issue of
how such ISI can be prevented by a better choice of pulse shape. Yes, there are
good pulse shapes that are wider than T .
Exercise 11.1. Modify pulsespec.m to reproduce Figures 11.4 and 11.5 for the
double-width pulse shape.
Exercise 11.2. Modify pulsespec.m to examine what happens when Hamming
pulse shapes of width 4T , 6T , and 10T are used. What is the bandwidth of the
resulting transmitted signals? Do you think it is possible to recover the message
from the received signals? Explain.

Output of pulse-shaping filter

Pulse Shaping and Receive Filtering

231

10

15

20

25

Spectrum of the output

Symbols
104
102
100

102

0.1

0.2

0.3 0.4 0.5 0.6 0.7

Normalized frequency

0.8

0.9

Figure 11.5 The top plot shows a segment of the output x of the pulse-shaping lter.
With this 3T -wide pulse shape, the pulses from adjacent symbols interfere with each
other. The bottom plot shows the magnitude spectrum of the output, which has the
same general contour as the spectrum of a single copy of the pulse, as in the bottom
plot of Figure 11.4.

11.3

Eye Diagrams
While the dierences between the pulse-shaped sequences in Figures 11.3 and
11.5 are apparent, it is dicult to see directly whether the distortions are serious; that is, whether they cause errors in the reconstructed data (i.e., the hard
decisions) at the receiver. After all, if the reconstructed message is the same as
the real message, then no harm has been done, even if the values of the received
analog waveform are not identical. This section uses a visualization tool called
eye diagrams that show how much smearing there is in the system, and whether
symbol errors will occur. Eye diagrams were encountered briey in Chapter 9
(refer back to Figure 9.8) when visualizing how the performance of the idealized
system degraded when various impairments were added.
Imagine an oscilloscope that traces out the received signal, with the special
feature that it is set to retrigger or restart the trace every nT seconds without
erasing the screen. Thus the horizontal axis of an eye diagram is the time over
which n symbols arrive, and the vertical axis is the value of the received waveform. In the ideal case, the trace begins with n pulses, each of which is a scaled
copy of p(t). Then the (n + 1)st to 2nth pulses arrive, and overlay the rst n,
though each is scaled according to its symbol value. When there is noise, channel
distortion, and timing jitter, the overlays will dier.

232

Chapter 11. Pulse Shaping and Receive Filtering

Optimum sampling times

Sensitivity to
timing error
3
Distortion
at zero
crossings

Figure 11.6 Interpreting

eye diagrams: a T -wide

Hamming blip is used to
pulse shape a 4-PAM
data sequence.

1
0
1
2
3

Noise
margin

The "eye"

kT

(k + 1)T

As the number of superimposed traces increases, the eye diagram becomes

denser, and gives a picture of how the pulse shape, channel, and other factors
combine to determine the reliability of the recovered message. Consider the n =
2-symbol eye diagram shown in Figure 11.6. In this gure, the message is taken
from the 4-PAM alphabet 1 3, and the Hamming pulse shape is used. The
center of the eye gives the best times to sample, since the openings (i.e., the
dierence between the received pulse shape when the data value is 1 and the
received pulse shape when the data value is 1, or between the received pulse
shape when the data value is 1 and the received pulse shape when the data value
is 3) are the largest. The width marked sensitivity to timing error shows the
range of time over which the samples quantize correctly. The noise margin is the
smallest vertical distance between the bands, and is proportional to the amount
of additive noise that can be resisted by the system without reporting erroneous
values.
Thus, eye diagrams such as Figure 11.6 give a clear picture of how good (or
how bad) a pulse shape may be. Sometimes the smearing in this gure is so great
that the open segment in the center disappears. The eye is said to be closed, and
this indicates that a simple quantizer (slicer) decision device will make mistakes
in recovering the data stream. This is not good!
For example, reconsider the 4-PAM example of the previous section that used
a triple-width Hamming pulse shape. The eye diagram is shown in Figure 11.7.
No noise was added when drawing this picture. In the left two plots there are
clear regions about the symbol locations where the eye is open. Samples taken
in these regions will be quantized correctly, though there are also regions where
mistakes will occur. The other plots show the closed eye diagrams using 3T -

233

Pulse Shaping and Receive Filtering

Eye diagram for the T-wide Hamming pulse shape

Eye diagram for the 3T-wide Hamming pulse shape

2
0

4
Eye diagram for the 2T-wide Hamming pulse shape

Eye diagram for the 5T-wide Hamming pulse shape

2
0

0
2
4

Symbols

Symbols

Figure 11.7 Eye diagrams for T -, 2T -, 3T -, and 5T -wide Hamming pulse shapes show
how the sensitivity to noises and timing errors increases as the pulse shape widens.
The closed eye in the bottom plot on the right means that symbol errors are
inevitable.

wide and 5T -wide Hamming pulse shapes. Symbol errors will inevitably occur,
even if all else in the system is ideal. All of the measures (the noise margin, the
sensitivity to timing, and the distortion at zero crossings) become progressively
worse, and ever smaller amounts of noise can cause decision errors.
The following code draws eye diagrams for the pulse shapes dened by the
variable ps. As in the pulse-shaping programs of the previous section, the N
binary data points are oversampled by a factor of M and the convolution of
the pulse shapes with the data uses the filter command. The reshape(x,a,b)
command changes a vector x of size a*b into a matrix with a rows and b columns,
which is used to segment x into b overlays, each a samples long. This works
smoothly with the Matlab plot function.
Listing 11.2. eyediag.m plot eye diagrams for a pulse shape
N=1000; m=pam (N, 2 , 1 ) ;
M=20; mup=zeros ( 1 ,N*M) ;
mup ( 1 :M:N*M)=m;
ps=hamming (M) ;
x=f i l t e r ( ps , 1 , mup ) ;
neye =5;
c=f l o o r ( length ( x ) / ( neye *M) ) ;
xp=x (N*Mneye *M* c +1:N*M) ;
plot ( reshape ( xp , neye *M, c ) )

%
%
%
%
%
%
%
%
%

random s i g n a l o f l e n g t h N
oversampling f a c t o r of M
o v e r s a m p l e by M
hamming p u l s e o f width M
c o n v o l v e p u l s e shape
s i z e of groups
number o f e y e s to p l o t
i gnor e t r a n s i e n t s at s t a r t
p l o t i n groups

Typical output of eyediag.m is shown in Figure 11.8. The rectangular pulse

shape in the top plot uses ps=ones(1,M), the Hamming pulse shape in the middle
uses ps=hamming(M), and the bottom plot uses a truncated sinc pulse shape
ps=srrc(L,0,M) for L=10 that is normalized so that the largest value is one. The
rectangular pulse is insensitive to timing errors, since sampling almost anywhere
(except right at the transition boundaries) will return the correct values. The
Hamming pulse shape has a wide eye, but may suer from a loss of SNR if the

234

Chapter 11. Pulse Shaping and Receive Filtering

Eye diagram for rectangular pulse shape

1
0
1
Eye diagram for Hamming pulse shape
1
Figure 11.8 Eye diagrams

for rectangular, Hamming,

and sinc pulse shapes with
binary data.

1
Eye diagram for sinc pulse shape
2
0
2

Optimum sampling times

samples are taken far from the center of the eye. Of the three, the sinc pulse is the
most sensitive, since it must be sampled near the correct instants or erroneous
values will result.
Exercise 11.3. Modify eyediag.m so that the data sequence is drawn from the
alphabet 1, 3, 5. Draw the appropriate eye diagram for the rectangular,
Hamming, and sinc pulse shapes.
Exercise 11.4. Modify eyediag.m to add noise to the pulse-shaped signal x. Use
the Matlab command v*randn for dierent values of v. Draw the appropriate
eye diagrams. For each pulse shape, how large can v be and still have the eye
remain open?
Exercise 11.5. Combine the previous two exercises. Modify eyediag.m as in
Exercise 11.3 so that the data sequence is drawn from the alphabet 1, 3, 5.
Add noise, and answer the same question as in Exercise 11.4. Which alphabet is
more susceptible to noise?
Exercise 11.6. TRUE or FALSE: For two rectangular impulse responses with
the same maximum magnitude but dierent time widths with T1 > T2 , the halfpower bandwidth of the frequency response of the pulse with width T1 exceeds
that of the pulse with width T2 .

Pulse Shaping and Receive Filtering

m[k] pulseshape
filter

channel

receive
filter

u[k]

y[k]
equalizer

kT+

235

Figure 11.9 The baseband

communications system of
Exercise 11.9.

Exercise 11.7. TRUE or FALSE: For the PAM baseband signals created by
a rectangular pulse and a triangular pulse with the same time width and the
same maximum amplitude, the half-power bandwidth of the sequence using the
triangular pulse exceeds that of the rectangular pulse.
Exercise 11.8. Exercise 2.10 asked the following question: TRUE or FALSE:
The atter the top of the pulse shape, the less sensitive the receiver is to small
timing osets. In the absence of noise and without matched ltering, this is
TRUE. Describe a noisy situation and a matched ltering that might cause the
answer to be FALSE.
Exercise 11.9. Consider the baseband communication system in Figure 11.9.
The dierence equation relating the symbols m[k] to the T -spaced equalizer input
u[k] for the chosen baud-timing factor  is
u[k] = 0.04m[k ] + 1.00m[k 1 ] + 0.60m[k 2 ] + 0.38m[k 3 ],
where is a nonnegative integer. The nite-impulse-response equalizer (lter) is
described by the dierence equation
y[k] = u[k] + u[k 1].
a. Suppose = 0.4 and the message source is binary 1. Is the system from
the source symbols m[k] to the equalizer output y[k] open-eye? Justify your
answer.
b. If the message source is 4-PAM (1, 3), can the system from m[k] to the
equalizer output y[k] be made open-eye by selection of ? If so, provide a
successful value of . If not, explain.
It is now easy to experiment with various pulse shapes. pulseshape2.m applies
a sinc-shaped pulse to a random binary sequence. Since the sinc pulse extends
innitely in time (both backward and forward), it cannot be represented exactly
in the computer (or in a real communication system) and the parameter L species the duration of the sinc, in terms of the number of symbol periods.
Listing 11.3. pulseshape2.m pulse shape a (random) sequence
N=2000; m=pam (N, 2 , 1 ) ;
M=10; mup=zeros ( 1 ,N*M) ;
mup ( 1 :M:N*M)=m;
L=10; ps=s r r c ( L , 0 ,M) ;
s c=sum( ps )/M;
x=f i l t e r ( ps / sc , 1 , mup ) ;

%
%
%
%
%
%

2PAM s i g n a l o f l e n g t h N
oversampling f a c t o r
o v e r s a m p l e by M
s i n c p u l s e shape 2 * L symbols wide
normalizing constant
c o n v o l v e p u l s e shape with data

236

Chapter 11. Pulse Shaping and Receive Filtering

Using a sinc pulse shape

0.4
0.3

0.2

0.1

0.1
10 8

2 0
2
4
Symbol interval

Pulse-shaped data sequence

10

8 10 12 14
Symbol number

16

18

20

Figure 11.10 A binary 1 data sequence is pulse shaped using a sinc pulse.

Figure 11.10 plots the output of pulseshape2.m. The left gure shows the
pulse shape while the right plot shows the analog pulse-shaped signal x(t)
over a duration of about 25 symbols. The function srrc.m rst appeared in the
discussion of interpolation in Section 6.4 (and again in Exercise 6.22), and is used
here to generate the sinc pulse shape. The sinc function that srrc.m produces
is actually scaled, and this eect is removed by normalizing with the variable
sc. Changing the second input argument from beta=0 to other small positive
numbers changes the shape of the curve, each with a sinc-like shape called a
square-root raised cosine. This will be discussed in greater detail in Sections 11.4
and 11.6. Typing help srrc in Matlab gives useful information on using the
function.
Observe that, even though the signal oscillates above and below the 1 lines,
there is no intersymbol interference. When using the Hamming pulse as in Figure 11.3, each binary value was clearly delineated. With the sinc pulse of Figure
11.10, the analog waveform is more complicated. However, at the correct sampling instances, it always returns to 1 (the horizontal lines at 1 are drawn to
help focus the eye on the crossing times). Unlike the T -wide Hamming shape,
the signal need not return to near zero with each symbol.
Exercise 11.10. In pulseshape2.m, examine the eect of using dierent oversampling rates M. Try M = 1, 5, 100.
Exercise 11.11. Change pulseshape2.m so that the data sequence is drawn
from the alphabet 1, 3, 5. Can you visually identify the correct values in the
pulse-shaped signal?
Exercise 11.12. In pulseshape2.m, examine the eect of using sinc approximations of dierent lengths L. Try L = 1, 5, 100, 1000.
Exercise 11.13. In pulseshape2.m, examine the eect of adding noise to the
received signal x. Try Matlab commands randn and rand. How large can the
noise be and still allow the data to be recognizable?

Pulse Shaping and Receive Filtering

237

Exercise 11.14. The goal is to design a frequency-division multiplexed (FDM)

system with a square-root raised cosine as the transmitter pulse shape. The
symbol period is T = 137 ms. The design uses T /4 sampling, pulse lengths of
8T , and a rollo factor of 0.9, but it does not work, since only three modulated
carrier signals t into the allotted bandwidth without multiuser interference.
Five are needed. What parameters in the design would you change and why?
Exercise 11.15. Using the code from Exercise 11.11, examine the eects of
adding noise in pulseshape2.m. Does the same amount of noise in the six-level
data have more or less eect than in the two-level data?
Exercise 11.16. Modify pulseshape2.m to include the eect of a nonunity
channel. Try both a highpass channel and a bandpass channel. Which appears
worse? What are reasonable criteria for better and worse in this context?
Exercise 11.17. A Matlab question: in pulseshape2.m, examine the eect of
using the filtfilt command for the convolution instead of the filter command. Can you gure out why the results are dierent?
Exercise 11.18. Another Matlab question: in pulseshape2.m, examine the
eect of using the conv command for the convolution instead of the filter
command. Can you gure out how to make this work?

11.4

Nyquist Pulses
Consider a multilevel signal drawn from a nite alphabet with values w(kT ),
where T is the sampling interval. Let p(t) be the impulse response of the linear
lter representing the pulse shape. The signal just after pulse shaping is
x(t) = wa (t) p(t),
where wa (t) is the pulse-train signal (11.1).
The corresponding output of the received lter is
y(t) = wa (t) p(t) hc (t) hR (t),
as depicted in Figure 11.1, where hc (t) is the impulse response of the channel and
hR (t) is the impulse response of the receive lter. Let hequiv (t) = p(t) hc (t)
hR (t) be the overall equivalent impulse response. Then the equivalent overall
frequency response (i.e., F {hequiv (t)}) is
Hequiv (f ) = P (f )Hc (f )HR (f ).

(11.2)

One approach would be to attempt to choose HR (f ) so that Hequiv (f ) attained

a desired value (such as a pure delay) for all f . This would be a specication of

238

Chapter 11. Pulse Shaping and Receive Filtering

the impulse response hequiv (t) at all t, since the Fourier transform is invertible.
But such a distortionless response is unnecessary, since it does not really matter
what happens between samples, but only what happens at the sample instants. In
other words, as long as the eye is open, the transmitted symbols are recoverable
by sampling at the correct times. In general, if the pulse shape is zero at all
integer multiples of kT but one, then it can have any shape in between without
causing intersymbol interference.
The condition that one pulse does not interfere with other pulses at subsequent
T -spaced sample instants is formalized by saying that hN Y Q (t) is a Nyquist pulse
if there is a such that

c k=0
(11.3)
hN Y Q (kT + ) =
0 k = 0
for all integers k, where c is some nonzero constant. The timing oset in (11.3)
will need to be found by the receiver.
A rectangular pulse with time-width less than T certainly satises (11.3), as
does any pulse shape that is less than T wide. But the bandwidth of the rectangular pulse (and other narrow pulse shapes such as the Hamming pulse shape)
may be too wide. Narrow pulse shapes do not utilize the spectrum eciently. But
if just any wide shape is used (such as the multiple-T -wide Hamming pulses),
then the eye may close. What is needed is a signal that is wide in time (and
narrow in frequency) that also fullls the Nyquist condition (11.3).
One possibility is the sinc pulse
hsinc (t) =

sin(f0 t)
,
f0 t

with f0 = 1/T . This has the narrowest possible spectrum, since it forms a rectangle in frequency (i.e., the frequency response of a lowpass lter). Assuming
that the clocks at the transmitter and receiver are synchronized so that = 0,
the sinc pulse is Nyquist because hsinc (0) = 1 and
sin(k)
=0
k
for all integers k =
 0. But there are several problems with the sinc pulse.
hsinc (kT ) =

r It has innite duration. In any real implementation, the pulse must be truncated.

r It is noncausal. In any real implementation, the truncated pulse must be

delayed.

r The steep band edges of the rectangular frequency function Hsinc (f ) are difcult to approximate.

r The sinc function sin(t)/t decays slowly, at a rate proportional to 1/t.

The slow decay (recall the plot of the sinc function in Figure 2.11 on page 32)
means that samples that are far apart in time can interact with each other when
there are even modest clock-synchronization errors.

Pulse Shaping and Receive Filtering

239

Fortunately, it is not necessary to choose between a pulse shape that is constrained to lie within a single symbol period T and the slowly decaying sinc.
While the sinc has the smallest dispersion in frequency, there are other pulse
shapes that are narrower in time and yet are only a little wider in frequency.
Trading o time and frequency behaviors can be tricky. Desirable pulse shapes
1. have appropriate zero crossings (i.e., they are Nyquist pulses),
2. have sloped band edges in the frequency domain, and
3. decay more rapidly in the time domain (compared with the sinc), while maintaining a narrow prole in the frequency domain.
One popular option is called the raised cosine-rollo (or raised cosine) lter.
It is dened by its Fourier transform

|f | < f1
1
HRC (f ) = 12 + 12 cos ((|f | f1 )/(2f )) f1 < |f | < B

0
|f | > B
where

r
r
r
r

B is the absolute bandwidth,

f0 is the 6 dB bandwidth, equal to 1/(2T ), half the symbol rate,
f = B f0 , and
f1 = f 0 f .

The corresponding time-domain function is




sin(2f0 t)
cos(2f t)
.
hRC (t) = F 1 {HRC (f )} = 2f0
2f0 t
1 (4f t)2

(11.4)

Dene the rollo factor = f /f0 . Figure 11.11 shows the magnitude spectrum
HRC (f ) of the raised cosine lter on the right and the associated time response
hRC (t) on the left, for a variety of rollo factors. With T = 1/(2f0), hRC (kT ) has
a factor sin(k)/(k) that is zero for all integer k = 0. Hence the raised cosine
is a Nyquist pulse. In fact, as 0, hRC (t) becomes a sinc.
The raised cosine pulse hRC (t) with nonzero has the following characteristics:

r zero crossings at desired times,

r band edges of HRC (f ) that are less severe than with a sinc pulse,
r an envelope that falls o at approximately 1/|t|3 for large t (look at (11.4)).
This is signicantly faster than 1/|t|. As the rollo factor increases from 0
to 1, the signicant part of the impulse response gets shorter.
Thus, we have seen several examples of Nyquist pulses: rectangular, Hamming,
sinc, and raised cosine with a variety of rollo factors. What is the general
principle that distinguishes Nyquist pulses from all others? A necessary and
sucient condition for a signal v(t) with Fourier transform V (f ) to be a Nyquist
pulse is that the sum (over all n) of V (f nf0 ) be constant. To see this, use the

240

Chapter 11. Pulse Shaping and Receive Filtering

1
=0

=1
0.5

= 0.5

=0

=1

0
= 0.5
0.5

3T

2T

2T

3T

f0/2

f0

Time

3f0/2

2f0

Frequency

Figure 11.11 Raised cosine pulse shape in the time and frequency domains.

sifting property of an impulse (A.56) to factor V (f ) from the sum:

$
%



V (f nf0 ) = V (f )
(f nf0 ) .
n=

n=

Given that convolution in the frequency domain is multiplication in the time

domain (A.40), on applying the denition of the Fourier transform, and using
the transform pair (from (A.28) with w(t) = 1 and W (f ) = (f ))
F{

(t kT )} =

k=

where f0 = 1/T , this becomes




V (f nf0 ) =

&

v(t) T

t=

n=

1 
(f nf0 ),
T n=

'%
(t kT )

ej2f t dt

k=

T v(kT )ej2f kT .

(11.5)

k=

If v(t) is a Nyquist pulse, the only nonzero term in the sum is v(0), and

V (f nf0 ) = T v(0).

n=

Thus, the sum of the V (f nf0 ) is a constant if v(t) is a Nyquist pulse. Conversely, if the sum of the V (f nf0 ) is a constant, then only the DC term in
(11.5) can be nonzero, and so v(t) is a Nyquist pulse.
Exercise 11.19. Write a Matlab routine that implements the raised cosine
impulse response (11.4) with rollo parameter . Hint 1: if you have trouble
with divide by zero errors, imitate the code in srrc.m. Plot the output of
your program for a variety of . Hint 2: there is an easy way to use the function
srrc.m.

Pulse Shaping and Receive Filtering

241

p(t)
3
1.5
0.2

0.6

Figure 11.12 Pulse shape for Exercise

1.0

11.23.
t (ms)

1.5
3

Exercise 11.20. Use your code from the previous exercise, along with
pulseshape2.m to apply raised cosine pulse shaping to a random binary
sequence. Can you spot the appropriate times to sample by eye?
Exercise 11.21. Use the code from the previous exercise and eyediag.m to draw
eye diagrams for the raised cosine pulse with rollo parameters r = 0, 0.5, 0.9,
1.0, 5.0. Compare these with the eye diagrams for rectangular and sinc functions.
Consider
a.
b.
c.
d.

sensitivity to timing errors,

peak distortion,
distortion of zero crossings, and
the noise margin.

Exercise 11.22. TRUE or FALSE: The impulse response of a series combination

of any -second-wide pulse-shape lter and its matched lter form a Nyquist
pulse shape for a T -spaced symbol sequence for any T .
Exercise 11.23. Consider the pulse shape p(t) of width 1.2 ms shown in Figure
11.12.
a. Is p(t) a Nyquist pulse for the symbol period T = 0.35 ms? Justify your
answer.
b. Is p(t) a Nyquist pulse for the symbol period T = 0.70 ms? Justify your
answer.
Exercise 11.24. Neither s1 (t) nor s2 (t) is a Nyquist pulse.
a. Can the product s1 (t)s2 (t) be a Nyquist pulse? Explain.
b. Can the convolution s1 (t) s2 (t) be a Nyquist pulse? Explain.
Intersymbol interference occurs when data values at one sample instant interfere with the data values at another sampling instant. Using Nyquist shapes such
as the rectangle, sinc, and raised cosine pulses removes the interference, at least
at the correct sampling instants, when the channel is ideal. The next sections

242

Chapter 11. Pulse Shaping and Receive Filtering

parlay this discussion of isolated pulse shapes into usable designs for the pulse
shaping and receive lters.

11.5

Matched Filtering
Communication systems must be robust against the presence of noises and other
disturbances that arise in the channel and in the various stages of processing.
Matched ltering is aimed at reducing the sensitivity to noise, which can be
specied in terms of the power spectral density (this is reviewed in some detail
in Appendix E).
Consider the ltering problem in which a message signal is added to a noise
signal and then both are passed through a linear lter. This occurs, for instance,
when the signal g(t) of Figure 11.1 is the output of the pulse-shaping lter
(i.e., no interferers are present), the channel is the identity, and there is noise
n(t) present. Assume that the noise is white; that is, assume that its power
spectral density Pn (f ) is equal to some constant for all frequencies.
The output y(t) of the linear lter with impulse response hR (t) can be
described as the superposition of two components, one driven by g(t) and the
other by n(t); that is,
y(t) = v(t) + w(t),
where
v(t) = hR (t) g(t) and w(t) = hR (t) n(t).
This is shown in block-diagram form in Figure 11.13. In both components, the
processing and the output signal are the same. The bottom diagram separates
out the component due to the signal (v(kT ), which contains the message ltered
through the pulse shape and the receive lter), and the component due to the
noise (w(kT ), which is the noise ltered through the receive lter). The goal of
this section is to nd the receive lter that maximizes the ratio of the power in
the signal v(kT ) to the power in the noise w(kT ) at the sample instants.
Consider choosing hR (t) so as to maximize the power of the signal v(t) at
time t = compared with the power in w(t) (i.e., to maximize v2 ( ) relative
to the total power of the noise component w(t)). This choice of hR (t) tends to
emphasize the signal v(t) and suppress the noise w(t). The argument proceeds
by nding the transfer function HR (f ) that corresponds to this hR (t).
From (E.2), the total power in w(t) is

Pw (f )df.
Pw =

From the inverse Fourier transform,


V (f )ej2f df,
v( ) =

243

Pulse Shaping and Receive Filtering

n(t)
m(kT)
g(t)
P( f )

y(t)

Pulse
shaping

Receive
filter

n(t)
HR( f )

Downsample

w(t)

Receive
filter
m(kT)

y(kT)

HR( f )

w(kT)
Downsample

v(t)

g(t)
P( f )

HR( f )

Pulse
shaping

Receive
filter

v(kT)

y(kT)

Downsample

Figure 11.13 The two block diagrams result in the same output. The top shows the

data ow in a normal implementation of pulse shaping and receive ltering. The

bottom shows an equivalent that allows easy comparison between the parts of the
output due to the signal (i.e., v(kT )) and the parts due to the noise (i.e., w(kT )).

where V (f ) = HR (f )G(f ). Thus,


2


HR (f )G(f )ej2f df  .
v 2 ( ) = 

Recall (E.3), which says that, for Y (f ) = HR (f )U (f ), Py (f ) = |HR (f )|2 Pu (f ).

Thus,
Pw (f ) = |HR (f )|2 Pn (f ) = |HR (f )|2 .
The quantity to be maximized can now be described by
(
| HR (f )G(f )ej2f df |2
v2 ( )
(
.
=
Pw
|HR (f )|2 df

Schwarzs inequality (A.57) says that


2 



2


a(x)b(x)dx
|a(x)| dx


(11.6)

|b(x)| dx ,
2

and equality occurs only when a(x) = kb (x). This converts (11.6) to
(

2
j2f 2
| df
v 2 ( )
|HR (f )| df
|G(f )e
(
,

Pw
|HR (f )|2 df
which is maximized with equality when
HR (f ) = k(G(f )ej2f ) .

(11.7)

244

Chapter 11. Pulse Shaping and Receive Filtering

HR (f ) must now be transformed to nd the corresponding impulse response

hR (t). Since Y (f ) = X(f ) when y(t) = x(t) (use the frequency-scaling property of Fourier transforms (A.38) with a scale factor of 1),
F 1 {W (f )} = w (t) F 1 {W (f )} = w (t).
Applying the time-shift property (A.37) yields
F 1 {W (f )ej2f Td } = w(t Td ).
Combining these two transform pairs yields
F 1 {(W (f )ej2f Td ) } = w ((t Td )) = w (Td t).
Thus, when g(t) is real,
F 1 {k(G(f )ej2f ) } = kg ( t) = kg( t).
Observe that this lter has the following characteristics.

r This lter results in the maximum signal-to-noise ratio of v 2 (t)/Pw at the

time instant t = for a noise signal with a at power spectral density.

r Because the impulse response of this lter is a scaled time reversal of the
r
r
r

pulse shape p(t), it is said to be matched to the pulse shape, and is called
a matched lter.
The shape of the magnitude spectrum of the matched lter HR (f ) is the same
as the magnitude spectrum G(f ).
The shape of the magnitude spectrum of G(f ) is the same as the shape of
the frequency response of the pulse shape P (f ) for a broadband m(kT ), as in
Section 11.1.
The matched lter for any lter with an even symmetric (about some t) timelimited impulse response is a delayed replica of that lter. The minimum delay
is the upper limit of the time-limited range of the impulse response.

The following code allows hands-on exploration of this theoretical result.

The pulse shape is dened by the variable ps (the default is the sinc function
srrc(L,0,M) for L=10). The receive lter is analogously dened by recfilt. As
usual, the symbol alphabet is easily specied by the pam subroutine, and the
system operates at an oversampling rate M. The noise is specied in n, and the
ratio of the powers is output as powv/poww. Observe that, for any pulse shape,
the ratio of the powers is maximized when the receive lter is the same as the
pulse shape (the fliplr command carries out the time reversal). This holds no
matter what the noise, no matter what the symbol alphabet, and no matter what
the pulse shape.
Listing 11.4. matchfilt.m test of SNR maximization
N=215; m=pam (N, 2 , 1 ) ;
M=10; mup=zeros ( 1 ,N*M) ;
mup ( 1 :M:N*M)=m;
L=10; ps=s r r c (L , 0 ,M) ;

%
%
%
%

2PAM s i g n a l o f l e n g t h N
oversampling f a c t o r
o v e r s a m p l e by M
d e f i n e p u l s e shape

Pulse Shaping and Receive Filtering

ps=ps / sqrt (sum( ps . 2 ) ) ;

n =0.5 * randn ( s i z e (mup ) ) ;
g=f i l t e r ( ps , 1 , mup ) ;
r e c f i l t =s r r c (L , 0 ,M) ;
r e c f i l t = r e c f i l t / sqrt (sum( r e c f i l t . 2 ) ) ;
v=f i l t e r ( f l i p l r ( r e c f i l t ) , 1 , g ) ;
w=f i l t e r ( f l i p l r ( r e c f i l t ) , 1 , n ) ;
vdownsamp=v ( 1 :M:N*M) ;
wdownsamp=w ( 1 :M:N*M) ;
powv=pow ( vdownsamp ) ;
poww=pow ( wdownsamp ) ;
powv/poww

%
%
%
%
%
%
%
%
%
%
%
%

245

and n o r m a l i z e
noise
c o n v o l v e ps with data
r e c e i v e f i l t e r H sub R
n o r m a l i z e p u l s e shape
matched f i l t e r with data
matched f i l t e r with n o i s e
downsample to symbol r a t e
downsample to symbol r a t e
power i n downsampled v
power i n downsampled w
ratio

In general, when the noise power spectral density is at (i.e., Pn (f ) = ), the

output of the matched lter may be realized by correlating the input to the
matched lter with the pulse shape p(t). To see this, recall that the output is
described by the convolution

s()h( )d
x() =

of the matched lter with the impulse response h(t). Given the pulse shape p(t)
and the assumption that the noise has at power spectral density, it follows that

p( t) 0 t T,
h(t) =
0
otherwise,
where is the delay used in the matched lter. Because h(t) is zero when t is
negative and when t > T , h( ) is zero for > and < T . Accordingly,
the limits on the integration can be converted to


s()p( ( ))d =
s()p()d.
x() =
=T

=T

This is the cross-correlation of p with s as dened in (8.3).

When Pn (f ) is not a constant, (11.6) becomes
(
| H(f )G(f )ej2f df |2
v2 ( )
(
.
=
2
Pw
Pn (f )|H(f )| df

To use the Schwarz inequality (A.57), associate a with H Pn and b with

Gej2f / Pn . Then (11.7) can be replaced by

(

2
j2f 2
| /Pn (f )df
v 2 ( )
|H(f )| Pn (f )df
|G(f )e
(
,

2
Pw
|H(f )| Pn (f )df
and equality occurs when a() = kb (); that is,
H(f ) =

kG (f )ej2f
.
Pn (f )

When the noise power spectral density Pn (f ) is not at, it shapes the matched
lter. Recall that the power spectral density of the noise can be computed from
its autocorrelation, as is shown in Appendix E.

246

Chapter 11. Pulse Shaping and Receive Filtering

w[k]
Figure 11.14 Baseband

m[k]

Transmit
filter

f [k]

Receive
filter
g[k]

y[k]

s[i]

k = iN +

communication system
considered in Exercise
11.29.

Exercise 11.25. Let the pulse shape be a T -wide Hamming blip. Use the code
in matchfilt.m to nd the ratio of the power in the downsampled v to that in
the downsampled w when
a. the receive lter is an SRRC with beta = 0, 0.1, 0.5,
b. the receive lter is a rectangular pulse, and
c. the receive lter is a 3T -wide Hamming pulse.
When is the ratio largest?
Exercise 11.26. Let the pulse shape be an SRRC with beta = 0.25. Use the
code in matchfilt.m to nd the ratio of the power in the downsampled v to that
in the downsampled w when
a. the receive lter is an SRRC with beta = 0, 0.1, 0.25, 0.5,
b. the receive lter is a rectangular pulse, and
c. the receive lter is a T -wide Hamming pulse.
When is the ratio largest?
Exercise 11.27. Let the symbol alphabet be 4-PAM.
a. Repeat Exercise 11.25.
b. Repeat Exercise 11.26.
Exercise 11.28. Create a noise sequence that is uniformly distributed (using
rand) with zero mean.
a. Repeat Exercise 11.25.
b. Repeat Exercise 11.26.
Exercise 11.29. Consider the baseband communication system in Figure 11.14.
The symbol period T is an integer multiple of the sample period T s , i.e., T = N Ts .
The message sequence is nonzero only each N th k, i.e., k = N Ts + Ts , where the
integer is within the range 0 < N and Ts is the on-sample baud-timing
oset at the transmitter.
a. Suppose that f [k] = {0.2, 0, 0, 1.0, 1.0, 0, 0, 0.2} for k = 0, 1, 2, 3, 4, 5, 6, 7 and
zero otherwise. Determine the causal g[k] that is the matched lter to f [k].
Arrange g[k] so that the nal nonzero element is as small as possible.

Pulse Shaping and Receive Filtering

247

b. For the g[k] specied in part (a), determine the smallest possible N so the path
from m[k] to y[k] forms an impulse response that would qualify as a Nyquist
pulse for some well-chosen baud-timing oset when y is downsampled by
N.
c. For the g[k] chosen in part (a) and the N chosen in part (b), determine the
downsampler oset for s[i] to be the nonzero entries of m[k] when w[k] = 0
for all k.

11.6

Matched Transmit and Receive Filters

While focusing separately on the pulse shaping and the receive ltering makes
sense pedagogically, the two are intimately tied together in the communication
system. This section notes that it is not really the pulse shape that should be
Nyquist, but rather the convolution of the pulse shape with the receive lter.
Recall the overall block diagram of the system in Figure 11.1, where it was
assumed that the portion of the system from upconversion (to passband) to nal
downconversion (back to baseband) is done perfectly and that the channel is just
the identity. Thus, the central portion of the system is eectively transparent
(except for the intrusion of noise). This simplies the system to the baseband
model in Figure 11.15.
The task is to design an appropriate pair of lters: a pulse shape for the
transmitter, and a receive lter that is matched to the pulse shape and the
presumed noise description. It is not crucial that the transmitted signal itself
have no intersymbol interference. Rather, the signal after the receive lter should
have no ISI. Thus, it is not the pulse shape that should satisfy the Nyquist pulse
condition, but the combination of the pulse shape and the receive lter.
The receive lter should simultaneously
1. allow no intersymbol interference at the receiver, and
2. maximize the signal-to-noise ratio.
Hence, it is the convolution of the pulse shape and the receive lter that should be
a Nyquist pulse, and the receive lter should be matched to the pulse shape. Considering candidate pulse shapes that are both symmetric and even about some
time t, the associated matched lter (modulo the associated delay) is the same as
the candidate pulse shape. What symmetric pulse shapes, when convolved with
themselves, form a Nyquist pulse? Previous sections examined several Nyquist
Noise
Figure 11.15 Noisy baseband

Signal

Pulse
shape

Receive
filter

communication system.

248

Chapter 11. Pulse Shaping and Receive Filtering

pulse shapes, the rectangle, the sinc, and the raised cosine. When convolved with
themselves, do any of these shapes remain Nyquist?
For a rectangle pulse shape and its rectangular matched lter, the convolution
is a triangle that is twice as wide as the original pulse shape. With precise timing,
(so that the sample occurs at the peak in the middle), this triangular pulse shape
is also a Nyquist pulse. This exact situation will be considered in detail in Section
12.2.
The convolution of a sinc function with itself is more easily viewed in the frequency domain as the point-by-point square of the transform. Since the transform
of the sinc is a rectangle, its square is a rectangle as well. The inverse transform
is consequently still a sinc, and is therefore a Nyquist pulse.
The raised cosine pulse fails. Its square in the frequency domain does not
retain the odd symmetry around the band edges, and the convolution of the
raised cosine with itself does not retain its original zero crossings. But the raised
cosine was the preferred Nyquist pulse because it conserves bandwidth eectively
and because its impulse response dies away quickly. One possibility is to dene
a new pulse shape that is the square root of the raised cosine (the square root is
taken in the frequency domain, not the time domain). This is called the squareroot raised cosine lter (SRRC). By denition, the square in frequency of the
SRRC (which is the raised cosine) is a Nyquist pulse.
The time-domain description of the SRRC pulse is found by taking the inverse
Fourier transform of the square root of the spectrum of the raised cosine pulse.
The answer is a bit complicated:

t=0
(1 + 4/)/ T

 
 

(+2)

sin 4
cos 4
+ (2)
t = 4
(11.8)
v(t) =
2T
2T

)+(4t/T ) cos((1+)t/T )
sin((1)t/T

otherwise.
T (t/T )(1(4t/T )2 )
Exercise 11.30. Plot the SRRC pulse in the time domain and show that it is
not a Nyquist pulse (because it does not cross zero at the desired times). The
Matlab routine srrc.m will make this easier.
Though the SRRC is not itself a Nyquist pulse, the convolution in time of two
SRRCs is a Nyquist pulse. The square-root raised cosine is the most commonly
used pulse in bandwidth-constrained communication systems.
Exercise 11.31. Consider the baseband communication system with a symbolscaled impulse train input
s(t) =

si (t iT )

i=

where T is the symbol period in seconds and 0 <  < 0.25 ms. The system contains a pulse-shaping lter P (f ), a channel transfer function C(f ) with additive

249

Pulse Shaping and Receive Filtering

x(t)

n(t)

1
s(t)
1

t (ms)

P( f )

C( f )

y[k]

y(t)
V( f )
kT +

Figure 11.16 The system of Exercise 11.31 and the time signal x(t) corresponding to
the inverse Fourier transform of X(f ).

noise n(t), and a receive lter V (f ), as shown in Figure 11.16. In addition, consider the time signal x(t) shown in the left part of Figure 11.16, where each of
the arcs is a half-circle. Let X(f ) be the Fourier transform of x(t).
a. If the inverse of P (f )C(f )V (f ) is X(f ), what is the highest symbol frequency
with no intersymbol interference supported by this communication system
when = 0?
b. If the inverse of P (f )C(f )V (f ) is X(f ), select a sampler time oset T /2 >
> T /2 to achieve the highest symbol frequency with no intersymbol interference. What is this symbol rate?
c. Consider designing the receive lter under the assumption that a symbol
period T can be (and will be) chosen so that samples yk of the receive lter
output suer no intersymbol interference. If the inverse of P (f )C(f ) is X(f ),
and the power spectral density of n(t) is a constant, plot the impulse response
of the causal, minimum-delay, matched receive lter for V (f ).

12 Timing Recovery

When the signal arrives at the receiver, it is a complicated analog waveform that
must be sampled in order to eventually recover the transmitted message. The
timing-oset experiments of Section 9.4.5 showed that one kind of stu that
can happen to the received signal is that the samples might inadvertently be
taken at inopportune moments. When this happens, the eye becomes closed
and the symbols are incorrectly decoded. Thus there needs to be a way to determine when to take the samples at the receiver. In accordance with the basic
system architecture of Chapter 2, this chapter focuses on baseband methods
of timing recovery (also called clock recovery). The problem is approached in
a familiar way: nd performance functions that have their maximum (or minimum) at the optimal point (i.e., at the correct sampling instants when the eye
is open widest). These performance functions are then used to dene adaptive
elements that iteratively estimate the sampling times. As usual, all other aspects
of the system are presumed to operate awlessly: the up and down conversions
are ideal, there are no interferers, and the channel is benign.
The discussion of timing recovery begins in Section 12.1 by showing how a
sampled version of the received signal x[k] can be written as a function of the
timing parameter , which dictates when to take samples. Section 12.2 gives
several examples that motivate several dierent possible performance functions,
(functions of x[k]), which lead to dierent methods of timing recovery. The
error between the received data values and the transmitted data (called the
source recovery error) is an obvious candidate, but it can be measured only
when the transmitted data are known or when there is an a priori known or
agreed-upon header (or training sequence). An alternative is to use the cluster
variance, which takes the square of the dierence between the received data
values and the nearest element of the source alphabet. This is analogous to
the decision-directed approach to carrier recovery (from Section 10.5), and an
adaptive element based on the cluster variance is derived and studied in Section
12.3. A popular alternative is to measure the power of the T -spaced output of
the matched lter. Maximizing this power (by choice of ) also leads to a good

Timing Recovery

251

answer, and an adaptive element based on output power maximization is detailed

in Section 12.4.
In order to understand the various performance functions, the error surfaces
are drawn. Interestingly, in many cases, the error surface for the cluster variance
has minima wherever the error surface for the output power has maxima. In these
cases, either method can be used as the basis for timing-recovery methods. On
the other hand, there are also situations when the error surfaces have extremal
points at dierent locations. In these cases, the error surface provides a simple
way of examining which method is most tting.

12.1

The Problem of Timing Recovery

The problem of timing recovery is to choose the instants at which to sample the
incoming (analog) signal. This can be translated into the mathematical problem
of nding a single parameter, the timing oset , which minimizes (or maximizes)
some function (such as the source recovery error, the cluster variance, or the
output power) of given the input. Clearly, the output of the sampler must
also be a function of , since species when the samples are taken. The rst
step is to write out exactly how the values of the samples depend on . Suppose
that the interval T between adjacent symbols is known exactly. Let gT (t) be
the pulse-shaping lter, gR (t) the receive lter, c(t) the impulse response of the
channel, s[i] the data from the signal alphabet, and w(t) the noise. Then the
baseband waveform at the input to the sampler can be written explicitly as
x(t) =

s[i](t iT ) gT (t) c(t) gR (t) + w(t) gR (t).

i=

Combining the three linear lters with

h(t) = gT (t) c(t) gR (t),

(12.1)

as shown in Figure 12.1, and sampling at interval T /M (M is again the oversampling factor) yields the sampled output at time kT /M + :






kT

s[i]h(t iT ) + w(t) gR (t)
.
+ =
x

M
i=
t=kT /M+

Assuming that the noise has the same distribution no matter when it is sampled, the noise term v[k] = w(t) gR (t)|t=kT /M+ is independent of . Thus, the
goal of the optimization is to nd so as to maximize or minimize some simple
function of the samples, such as






kT
kT
+ =
+ iT + v[k].
(12.2)
s[i]h
x[k] = x
M
M
i=

252

Chapter 12. Timing Recovery

h(t) = gR(t)*c(t)*gT(t)
Figure 12.1 The transfer

gT(t)

gR(t)

c(t)

s[i] Transmit
filter

Channel

Sampler
x(kT/M + )

Receive x(t)
filter

function h combines the

eects of the transmitter
pulse shaping gT , the
channel c, and the receive
lter gR .

w(t)

Sampler
(a)

ASP

DSP
Sampler
(b)

ASP

DSP

Sampler
(c)

ASP

DSP

Figure 12.2 Three generic structures for timing recovery. In (a), an analog processor

determines when the sampling instants will occur. In (b), a digital post-processor is
used to determine when to sample. In (c), the sampling instants are chosen by a
free-running clock, and digital post-processing is used to recover the values of the
received signal that would have occurred at the optimal sampling instants.

There are three ways in which timing-recovery algorithms can be implemented,

and these are shown in Figure 12.2. In the rst, an analog processor determines
when the sampling instants will occur. In the second, a digital post-processor
is used to determine when to sample. In the third, the sampling instants are
chosen by a free-running clock, and digital post-processing (interpolation) is
used to recover the values of the received signal that would have occurred at the
optimal sampling instants. The adaptive elements of the next sections can be
implemented in any of the three ways, though in digital radio systems the trend
is to remove as much of the calculation from analog circuitry as possible.

12.2

An Example
This section works out in complete and gory detail what may be the simplest
case of timing recovery. More realistic situations will be considered (by numerical
methods) in later sections.

Timing Recovery

1.0

Figure 12.3 For the example of this section,

the concatenation of the pulse shape, the
channel, and the receive ltering (h(t) of
(12.1)) is assumed to be a symmetric triangle
wave with unity amplitude and support 2T .

h(t)
0 0

T T + 0

253

2T

Time t

Consider a noise-free binary 1 baseband communication system in which the

transmitter and receiver have agreed on the rate of data ow (one symbol every
T seconds, with an oversampling factor of M = 1). The goal is to select the
instants kT + at which to sample, that is, to nd the oset . Suppose that
the pulse-shaping lter is chosen so that h(t) is Nyquist; that is,

1 k = 1,
h(kT ) =
0 k = 1.
The sampled output sequence is the amplitude-modulated impulse train s[i] convolved with a lter that is the concatenation of the pulse shaping, the channel,
and the receive ltering, and evaluated at the sampler closure times, as in (12.2).
Thus,




x[k] =
s[i]h(t iT )
.

i

t=kT +

To keep the computations tractable, suppose that h(t) has the triangular shape
shown in Figure 12.3. This might occur, for instance, if the pulse-shaping lter
and the receive lter are both rectangular pulses of width T and the channel is
the identity.
There are three cases to consider: = 0, > 0, and < 0.

r With = 0, which synchronizes the sampler to the transmitter pulse times,

h(t iT )|t=kT + = h(kT + iT )
= h((k i)T + ) = h((k i)T )

1 k i = 1 i = k 1,
=
0 otherwise.
In this case, x[k] = s[k 1] and the system is a pure delay.

r With = 0 > 0, the only two nonzero points among the sampled impulse
response are at h(0 ) and h(T + 0 ), as illustrated in Figure 12.3. Therefore,

1 0 /T k i = 1,
h(t iT )|t=kT +0 = h((k i)T + 0 ) = 0 /T
k i = 0,

0
otherwise.

254

Chapter 12. Timing Recovery

To work out a numerical example, let k = 6. Then


s[i]h((6 i)T + 0 )
x[6] =
i

= s[6]h(0 ) + s[5]h(T + 0 ) = s[6]


0 
0
+ s[5] 1
.
T
T

Since the data are binary, there are four possibilities for the pair (s[5], s[6]):
(s[5], s[6]) = (+1, +1) x[6] = 0 /T + 1 0 /T = 1,
(s[5], s[6]) = (+1, 1) x[6] = 0 /T + 1 0 /T = 1 20 /T,
(s[5], s[6]) = (1, +1) x[6] = 0 /T 1 + 0 /T = 1 + 20 /T,
(s[5], s[6]) = (1, 1) x[6] = 0 /T 1 + 0 /T = 1.

(12.3)

Note that two of the possibilities for x[6] give correct values for s[5], while
two are incorrect.
With = 0 < 0, the only two nonzero points among the sampled impulse
response are at h(2T + 0 ) and h(T + 0 ). In this case,

1 |0 |/T k i = 1,
h(t iT )|t=kT +0 = |0 |/T
k i = 2,

0
otherwise.

The next two examples look at two possible measures of the quality of : the
cluster variance and the output power.

Example 12.1. The decision device Q(x[k]) quantizes its argument to the nearest
member of the symbol alphabet. For binary data, this is the signum operator that
maps any positive number to +1 and any negative number to 1. If T /2 < 0 <
T /2, then Q(x[k]) = s[k 1] for all k, the eye is open, and the source recovery
error can be written as e[k] = s[k 1] x[k] = Q(x[k]) x[k]. Continuing the
example, and assuming that all symbol-pair choices are equally likely, the average
squared error at time k = 6 is

#
2 
2
 

2|
2|
1
|
|
0
0
(1 1)2 + 1 1
+ 1 1 +
+ (1 (1))2 ,
4
T
T
which simplies to avg{e2 [6]} = 202 /T 2 . The same result occurs for any other
k.
If 0 is outside the range of (T /2, T /2) then Q(x[k]) no longer equals s[k 1]
(but it does equal s[j] for some j = k 1). The cluster variance
CV = avg{e2 [k]} = avg{(Q(x[k]) x[k])2 }

(12.4)

is a useful measure, and this is plotted in Figure 12.4 as a function of . The

periodic nature of the function is clear, and the problem of timing recovery can
be viewed as a one-dimensional search for that minimizes the CV.

Timing Recovery

255

avg{(Q(x) x)2}
0.5

3T/2

T/2

Figure 12.4 Cluster variance as a

function of oset timing .

T/2

3T/2

Timing offset

avg{x2}
1.0
Figure 12.5 Average squared
output as a function of timing
oset .

0.5

3T/2

T/2

T/2

3T/2

Timing offset

Example 12.2. Another measure of the quality of the timing parameter is given
by the power (average energy) of the x[k]. Using the four formulas (12.3), and
observing that analogous formulas also apply when 0 < 0, the average energy
can be calculated for any k by
avg{x2 [k]} = (1/4)[(1)2 + (1 (2| |/T ))2 + (1 + (2| |/T ))2 + (1)2 ]
= (1/4)[2 + 2(1 (2| |/T ))2 ] = 1 (2| |/T ) + (2| |2 /T 2),
assuming that the four symbol pairs are equally likely. The average of x2 [k] is
plotted in Figure 12.5 as a function of . Over T /2 < < T /2, this average is
maximized with = 0. Thus, the problem of timing recovery can also be viewed
as a one-dimensional search for the that maximizes avg{x2 [k]}.
Thus, at least in the simple case of binary transmission with h(t) a triangular
pulse, the optimal timing oset (for the plots in Figures 12.4 and 12.5, at = nT
for integer n) can be obtained either by minimizing the cluster variance or by
maximizing the output power. In more general situations, the two measures
might not be optimized at the same point. Which approach is best when

r
r
r
r

there is channel noise,

the source alphabet is multilevel,
more common pulse shapes are used, and
there is intersymbol interference?

The next two sections show how to design adaptive elements that carry out
these minimizations and maximizations. The error surfaces corresponding to

256

Chapter 12. Timing Recovery

the performance functions will be used to gain insight into the behavior of the
methods even in nonideal situations.

12.3

Decision-Directed Timing Recovery

If the combination of the pulse shape, channel, and matched lter has the Nyquist
property, then the value of the waveform is exactly equal to the value of the data
at the correct sampling times. Thus, there is an obvious choice for the performance function: nd the sampling instants at which the dierences between
the received values and the transmitted values are smallest. This is called the
source recovery error and can be used when the transmitted data are knownfor
instance, when there is a training sequence. However, if the data are unavailable
(which is the normal situation), then the source recovery error cannot be measured and hence cannot form the basis of a timing-recovery algorithm.
The previous section suggested that a possible substitute is to use the cluster
variance avg{(Q(x[k]) x[k])2 }. Remember that the samples x[k] = x(kT /M +
) are functions of because species when the samples are taken, as is evident
from (12.2). Thus, the goal of the optimization is to nd so as to minimize
JCV ( ) = avg{(Q(x[k]) x[k])2 }.

(12.5)

Solving for directly is nontrivial, but JCV ( ) can be used as the basis for an
adaptive element

dJCV ( ) 
.
(12.6)
[k + 1] = [k]


d
= [k]
Using the approximation (G.12), which swaps the order of the derivative and the
average, yields

#
dJCV ( )
d (Q(x[k]) x[k])2
avg
d
d


dx[k]
= 2 avg (Q(x[k]) x[k])
.
(12.7)
d
The derivative of x[k] can be approximated numerically. One way of doing this
is to use
dx(kT /M + )
x(kT /M + + ) x(kT /M + )
dx[k]
=

,
(12.8)
d
d
2
which is valid for small . Substituting (12.8) and (12.7) into (12.6) and evaluating at = [k] gives the algorithm
[k + 1] = [k] + avg {(Q(x[k]) x[k])
[x (kT /M + [k] + ) x (kT /M + [k] )]} ,

Timing Recovery

Sampler
x(t)

x[k]

x(kT/M + [k])
Resample

257

Q( )
+

Resample

Resample

x(kT/M + [k] + )

x(kT/M + [k] )

+
[k]

Figure 12.6 One implementation of the adaptive element (12.9) uses three digital
interpolations (resamplers). After the [k] converge, the output x[k] is a sampled
version of the input x(t), with the samples taken at times that minimize the cluster
variance.

where the stepsize =

/. As usual, this algorithm acts as a lowpass lter to
smooth or average the estimates of , and it is common to remove the explicit
outer averaging operation from the update, which leads to
[k + 1] = [k] + (Q(x[k]) x[k])
[x (kT /M + [k] + ) x (kT /M + [k] )] .

(12.9)

If the [k] are too noisy, the stepsize can be decreased (or the length of the
average, if present, can be increased), although these will inevitably slow the
convergence of the algorithm.
The algorithm (12.9) is easy to implement, though it requires samples of the
waveform x(t) at three dierent points: x(kT /M + [k] ), x(kT /M + [k]),
and x(kT /M + [k] + ). One possibility is to straightforwardly sample three
times. Since sampling is done by hardware, this is a hardware-intensive solution. Alternatively, the values can be interpolated. Recall from the sampling
theorem that a waveform can be reconstructed exactly at any point, as long as
it is sampled faster than twice the highest frequency. This is useful since the
values at x(kT /M + [k] ) and at x(kT /M + [k] + ) can be interpolated
from the nearby samples x[k]. Recall that interpolation was discussed in Section 6.4, and the Matlab routine interpsinc.m on page 112 makes it easy to
implement bandlimited interpolation and reconstruction. Of course, this requires
extra calculations, so it is a more software-intensive solution. This strategy is
diagrammed in Figure 12.6.
The following code prepares the transmitted signal that will be used subsequently to simulate the timing-recovery methods. The user species the signal
constellation (the default is 4-PAM), the number of data points n, and the oversampling factor m. The channel is allowed to be nonunity, and a square-root raised
cosine pulse with width 2*l+1 and rollo beta is used as the default transmit
(pulse-shaping) lter. An initial timing oset is specied in toffset, and the

258

Chapter 12. Timing Recovery

code implements this delay with an oset in the srrc.m function. The matched
lter is implemented using the same SRRC (but without the time delay). Thus,
the timing oset is not known at the receiver.
Listing 12.1. clockrecDD.m (part 1) prepare transmitted signal
n =10000;
m=2;
beta = 0 . 3 ;
l =50;
chan = [ 1 ] ;
t o f f s e t = 0.3;
p u l s h a p=s r r c ( l , beta ,m, t o f f s e t ) ;
s=pam ( n , 4 , 5 ) ;
sup=zeros ( 1 , n *m) ;
sup ( 1 :m: n *m)= s ;
hh=conv ( pulshap , chan ) ;
r=conv ( hh , sup ) ;
m a t c h f i l t=s r r c ( l , beta ,m, 0 ) ;
x=conv ( r , m a t c h f i l t ) ;

%
%
%
%
%
%
%
%
%
%
%
%
%
%

number o f data p o i n t s
oversampling f a c t o r
r o l l o f f p a r a m eter f o r s r r c
1/2 l e n g t h o f p u l s e shape
T/m ch a n n e l
i n i t i a l timing o f f s e t
p u l s e shape with o f f s e t
random data s e q u e n c e v a r=5
upsample th e data , p l a c i n g . . .
m1 z e r o s between data p o i n t s
and p u l s e shape
to g e t r e c e i v e d s i g n a l
matched f i l t e r =s r r c
c o n v o l v e with matched f i l t e r

The goal of the timing recovery in clockrecDD.m is to nd (the negative of)

the value of toffset using only the received signalthat is, to have tau converge
to -toffset. The adaptive element is implemented in clockrecDD.m using the
iterative cluster variance algorithm (12.9). The algorithm is initialized with an
oset estimate of tau=0 and stepsize mu. The received signal is sampled at m
times the symbol rate, and the while loop runs though the data, incrementing i
once for each symbol (and incrementing tnow by m for each symbol). The osets
tau and tau+m are indistinguishable from the point of view of the algorithm. The
update term contains the interpolated value xs as well as two other interpolated
values to the left and right that are used to approximate the derivative term.
Listing 12.2. clockrecDD.m (part 2) clock recovery minimizing cluster variance
tnow=l *m+1; tau =0; x s=zeros ( 1 , n ) ;
% i n i t i a l i z e variables
t a u s a v e=zeros ( 1 , n ) ; t a u s a v e (1)= tau ; i =0;
mu= 0 . 0 1 ;
% algorithm s t e p s i z e
del ta =0.1;
% time f o r d e r i v a t i v e
while tnow<length ( x)2 * l *m
% run i t e r a t i o n
i=i +1;
x s ( i )= i n t e r p s i n c ( x , tnow+tau , l ) ;
% i n t e r p v a l u e a t tnow+tau
x d e l t a p=i n t e r p s i n c ( x , tnow+tau+d e l t a , l ) ; % v a l u e to r i g h t
x d e l ta m=i n t e r p s i n c ( x , tnow+taud e l t a , l ) ; % v a l u e to l e f t
% numerical d e r i v a t i v e
dx=x d e l t a p x d e l ta m ;
qx=q u a n t a l p h ( x s ( i ) , [ 3 , 1 , 1 , 3 ] ) ; % q u a n t i z e to a l p h a b e t
tau=tau+mu* dx * ( qxx s ( i ) ) ;
% a l g update : DD
tnow=tnow+m; t a u s a v e ( i )= tau ;
% save f o r p l o t t i n g
end

Typical output of the program is plotted in Figure 12.7, which shows the 4PAM constellation diagram, along with the trajectory of the oset estimation as
it converges towards the negative of the unknown value 0.3. Observe that,
initially, the values are widely dispersed about the required 4-PAM values, but,
as the algorithm nears its convergent point, the estimated values of the symbols
converge nicely.

Timing Recovery

259

Offset estimates

Symbol estimates

Constellation history
4
2
0
2
4

1000

2000

3000

4000

5000

1000

2000

3000
Iterations

4000

5000

0.3
2

Figure 12.7 Output of the

program clockrecDD.m shows
the symbol estimates in the
top plot and the trajectory of
the oset estimation in the
bottom.

0.1
0

As usual, a good way to conceptualize the action of the adaptive element is

to draw the error surfacein this case, to plot JCV ( ) of (12.5) as a function
of the timing oset . In the examples of Section 12.2, the error surface was
drawn by exhaustively writing down all the possible input sequences, and evaluating the performance function explicitly in terms of the oset . In the binary
setup with an identity channel, where the pulse shape is only 2T long, and with
M = 1 oversampling, there were only four cases to consider. But when the pulse
shape and channel are long and the constellation has many elements, the number of cases grows rapidly. Since this can get out of hand, an experimental
method can be used to approximate the error surface. For each timing oset,
the code in clockrecDDcost.m chooses n random input sequences, evaluates the
performance function, and averages.
Listing 12.3. clockrecDDcost.m error surfaces for cluster variance performance function
l =10;
beta = 0 . 5 ;
m=20;
ps=s r r c ( l , beta ,m) ;
p s r c=conv ( ps , ps ) ;
p s r c=p s r c ( l *m+1:3 * l *m+1);
c o s t=zeros ( 1 ,m) ; n =20000;
x=zeros ( 1 , n ) ;
f o r i =1:m
pt=p s r c ( i :m: end ) ;
f o r k =1:n
rd=pam ( length ( pt ) , 4 , 5 ) ;
x ( k)=sum( rd . * pt ) ;
end
e r r=q u a n ta l p h ( x , [ 3 , 1 , 1 , 3 ] ) x ;
c o s t ( i )=sum( e r r . 2 ) / length ( e r r ) ;
end

%
%
%
%
%
%
%

1/2 d u r a t i o n o f p u l s e shape
r o l l o f f f o r p u l s e shape
eval u ate at m d i f f e r e n t p oi nts
make s r r c p u l s e shape
c o n v o l v e 2 s r r c s to g e t r c
t r u n c a t e to same l e n g t h a s ps
e x p e r i m e n t a l p er f o r m a n ce

%
%
%
%
%

f o r each o f f s e t
r c i s s h i f t e d i /m o f a symbol
do i t n t i m e s
random 4PAM v e c t o r
r e c e i v e d data p o i n t w/ I S I

% q u a n t i z e to a l p h a b e t
% DD p e r f o r m a n c e f u n c t i o n

The output of clockrecDDcost.m is shown in Figure 12.8. The error surface

is plotted for the SRRC with ve dierent rollo factors. For all , the correct

Chapter 12. Timing Recovery

= 0.2
Value of performance function

260

0.5

=0

= 0.4
Figure 12.8 The performance

0.4

function (12.5) is plotted as a

function of the timing oset
for ve dierent pulse shapes
characterized by dierent rollo
factors . The correct answer is
at the global minimum at = 0.

0.3
= 0.6

0.2
0.1
0

= 0.8
0

T/2
Timing offset

answer at = 0 is a minimum. For small values of , this is the only minimum

and the error surface is unimodal over each period. In these cases, no matter
where is initialized, it should converge to the correct answer. As is increased,
however, the error surface attens across its top and gains two extra minima.
These represent erroneous values of to which the adaptive element may converge. Thus, the error surface can warn the system designer to expect certain
kinds of failure modes in certain situations (such as certain pulse shapes).
Exercise 12.1. Use clockrecDD.m to play with the clock-recovery algorithm.
a. How does mu aect the convergence rate? What range of stepsizes works?
b. How does the signal constellation of the input aect the convergent value of
tau? (Try 2-PAM and 6-PAM. Remember to quantize properly in the algorithm update.)
Exercise 12.2. Implement a rectangular pulse shape. Does this work better or
worse than the SRRC?
Exercise 12.3. Add noise to the signal (add a zero-mean noise to the received
signal using the Matlab randn function). How does this aect the convergence
of the timing-oset parameter tau. Does it change the nal converged value?
Exercise 12.4. Modify clockrecDD.m by setting toffset=-0.8. This starts the
iteration in a closed-eye situation. How many iterations does it take to open the
eye? What is the convergent value?
Exercise 12.5. Modify clockrecDD.m by changing the channel. How does this
aect the convergence speed of the algorithm? Do dierent channels change the
convergent value? Can you think of a way to predict (given a channel) what the
convergent value will be?

Timing Recovery

261

Exercise 12.6. Modify the algorithm (12.9) so that it minimizes the source
recovery error (s[k d] x[k])2 , where d is some (integer) delay. You will need to
assume that the message s[k] is known at the receiver. Implement the algorithm
by modifying the code in clockrecDDcost.m. Compare the new algorithm with
the old in terms of convergence speed and nal convergent value.
Exercise 12.7. Using the source recovery error algorithm of Exercise 12.6, examine the eect of dierent pulse shapes. Draw the error surfaces (mimic the code in
clockrecDDcost.m). What happens when you have the wrong d? What happens
when you have the right d?
Exercise 12.8. Investigate how the error surface depends on the input signal.
a. Draw the error surface for the DD timing-recovery algorithm when the inputs
are binary 1.
b. Draw the error surface when the inputs are drawn from the 4-PAM constellation, for the special case in which the symbol 3 never occurs.

12.4

Timing Recovery via Output Power Maximization

Any timing-recovery algorithm must choose the instants at which to sample the
received signal. The previous section showed that this can be translated into the
mathematical problem of nding a single parameter, the timing oset , which
minimizes the cluster variance. The extended example of Section 12.2 suggests
that maximizing the average of the received power (i.e., maximizing avg{x2 [k]})
leads to the same solutions as minimizing the cluster variance. Accordingly, this
section builds an element that adapts so as to nd the sampling instants at
which the power (in the sampled version of the received signal) is maximized.
To be precise, the goal of the optimization is to nd so as to maximize

 
kT
2
2
+
,
(12.10)
JOP ( ) = avg{x [k]} = avg x
M
which can be optimized using an adaptive element

dJOP ( ) 
[k + 1] = [k] +

.

d
= [k]

(12.11)

The updates proceed in the same direction as the gradient (rather than minus the
gradient) because the goal is to maximize, to nd the that leads to the largest
value of JOP ( ) (rather than the smallest). The derivative of JOP ( ) can be
approximated using (G.12) to swap the dierentiation and averaging operations
dJOP ( )
avg
d

dx2 [k]
d



dx[k]
= 2 avg x[k]
.
d

(12.12)

262

Chapter 12. Timing Recovery

Sampler
x(t)

Resample

Resample

x(kT/M + [k])

x[k]

x(kT/M + [k] + )

x(kT/M + [k] )
Resample

[k]

Figure 12.9 One implementation of the adaptive element (12.13) uses three digital

interpolations (resamplers). After the [k] converge, the output x[k] is a sampled
version of the input x(t), with the samples taken at times that maximize the power of
the output.

The derivative of x[k] can be approximated numerically. One way of doing this
is to use (12.8), which is valid for small . Substituting (12.8) and (12.12) into
(12.11) and evaluating at = [k] gives the algorithm

 



kT
kT
[k + 1] = [k] + avg x[k] x
+ [k] + x
+ [k]
,
M
M
where the stepsize =
/. As usual, the small-stepsize algorithm acts as a
lowpass lter to smooth the estimates of , and it is common to remove the
explicit outer averaging operation, leading to



 
kT
kT
+ [k] + x
+ [k] . (12.13)
[k + 1] = [k] + x[k] x
M
M
If [k] is noisy, then can be decreased (or the length of the average, if present,
can be increased), although these will inevitably slow the convergence of the
algorithm.
Using the algorithm (12.13) is similar to implementing the cluster variance
scheme (12.9), and a software-intensive solution is diagrammed in Figure
12.9. This uses interpolation (resampling) to reconstruct the values of x(t) at
x(kT /M + [k] ) and at x(kT /M + [k] + ) from nearby samples x[k]. As
suggested by Figure 12.2, the same idea can be implemented in analog, hybrid,
or digital form.
The following program implements the timing-recovery algorithm using
the recursive output-power-maximization algorithm (12.13). The user species the transmitted signal, channel, and pulse shaping exactly as in part 1 of
clockrecDD.m. An initial timing oset toffset is specied, and the algorithm
in clockrecOP.m tries to nd (the negative of) this value using only the received
signal.
Listing 12.4. clockrecOP.m clock recovery maximizing output power
tnow=l *m+1; tau =0; x s=zeros ( 1 , n ) ;

% initialize

variables

Offset estimates

Symbol estimates

Timing Recovery

263

Constellation history

1.5
1
0

Figure 12.10 Output of the

program clockrecOP.m shows
the estimates of the symbols
in the top plot and the
trajectory of the oset
estimates in the bottom.

1
0.4
0.2
0
0

1000

2000

3000
Iterations

4000

5000

t a u s a v e=zeros ( 1 , n ) ; t a u s a v e (1)= tau ; i =0;

mu= 0 . 0 5 ;
del ta =0.1;
while tnow<length ( x) l *m
i=i +1;
x s ( i )= i n t e r p s i n c ( x , tnow+tau , l ) ;
x d e l t a p=i n t e r p s i n c ( x , tnow+tau+d e l t a , l ) ;
x d e l ta m=i n t e r p s i n c ( x , tnow+tau d e l t a , l ) ;
dx=x d e l t a p x d e l ta m ;
tau=tau+mu* dx * x s ( i ) ;
tnow=tnow+m; t a u s a v e ( i )= tau ;
end

% algorithm s t e p s i z e
% time f o r d e r i v a t i v e
% run i t e r a t i o n
%
%
%
%
%
%

i n t e r p a t tnow+tau
v a l u e to r i g h t
v a l u e to l e f t
numerical d e r i v a t i v e
a l g update ( e n e r g y )
save f o r p l o t t i n g

Typical output of the program is plotted in Figure 12.10. For this plot, the
message was drawn from a 2-PAM binary signal, which is recovered nicely by
the algorithm, as shown in the top plot. The bottom plot shows the trajectory
of the oset estimation as it converges to the unknown value at -toffset.
The error surface for the output-power-maximization algorithm can be drawn
using the same experimental method as was used in clockrecDDcost.m.
Replacing the line that calculates the performance function with
c o s t ( i )=sum( x . 2 ) / length ( x ) ;

calculates the error surface for the output power algorithm (12.13). Figure 12.11
shows this, along with three variants:
1. the average value of the absolute value of the output of the sampler
avg{|x[k]|},
2. the average of the fourth power of the output of the sampler avg{x4 [k]}, and
3. the average of the dispersion avg{(x2 [k] 1)2 }.
Clearly, some of these require maximization (the output power and the absolute
value), while others require minimization (the fourth power and the dispersion).
While they all behave more or less analogously in this easy setting (the gure

Chapter 12. Timing Recovery

Fourth Power
Value of performance functions

264

1.2
Output Power

Figure 12.11 Four performance

functions that can be used for

timing recovery, plotted as a
function of the timing oset . In
this gure, the optimal answer is
at = 0. Some of the performance
functions must be minimized and
some must be maximized.

0.8
0.6

Abs. Value

Dispersion

0.4
0.2
0

T/2
Timing offset

shows the 2-PAM case with an SRRC pulse shape with beta=0.5), the maxima
(or minima) may occur at dierent values of in more extreme settings.
Exercise 12.9. TRUE or FALSE: The optimum settings of timing recovery via
output power maximization with and without intersymbol interference in the
analog channel are the same.
Exercise 12.10. Use the code in clockrecOP.m to play with the output power
clock-recovery algorithm. How does mu aect the convergence rate? What range
of stepsizes works? How does the signal constellation of the input aect the
convergent value of tau (try 4-PAM and 8-PAM)?
Exercise 12.11. Implement a rectangular pulse shape. Does this work better or
worse than the SRRC?
Exercise 12.12. Add noise to the signal (add a zero-mean noise to the received
signal using the Matlab randn function). How does this aect the convergence
of the timing-oset parameter tau? Does it change the nal converged value?
Exercise 12.13. Modify clockrecOP.m by setting toffset=-1. This starts the
iteration in a closed-eye situation. How many iterations does it take to open
the eye? What is the convergent value? Try other values of toffset. Can you
predict what the nal convergent value will be? Try toffset=-2.3. Now let the
oversampling factor be m = 4 and answer the same questions.
Exercise 12.14. Redo Figure 12.11 using a sinc pulse shape. What happens to
the output power performance function?
Exercise 12.15. Redo Figure 12.11 using a T -wide Hamming pulse shape.
Which of the four performance functions need to be minimized and which need
to be maximized?

Timing Recovery

265

Exercise 12.16. Consider the sampled version x(kT + ) of the baseband signal
x(t) recovered by the receiver in a 2-PAM communication system with source
alphabet 1. In ideal circumstances, the baud-timing variable could be set
so x(kT + ) exactly matched the source symbol sequence at the transmitter.
Suppose is selected as the value that optimizes
JDM ( ) =

1
N

k
0 +N

(1 x2 (kTs + ))2

k=k0 +1

for a suitably large N .

a. Should JDM be minimized or maximized?
b. Use the choice in part (a) and the approximation
x(kTs + ) x(kTs + )
d(x(kTs + ))

d

for small  > 0 to derive a gradient algorithm that optimizes JDM .
Exercise 12.17. Consider a resampler with input x(kTs ) and output x(kTs + ).
Use the approximation in Exercise 12.16(b) to derive an approximate gradientdescent algorithm that minimizes the fourth-power performance function
JF P ( ) =

1
N

k
0 +N

x4 (kTs + )

k=k0 +1

for a suitably large N .

Exercise 12.18. Implement the algorithm from Exercise 12.16 using
clockrecOP.m as a basis and compare the behavior with the output power maximization. Consider
a. convergence speed,
b. resilience against noise, and
c. misadjustment when ISI is present.
Now answer the same questions for the fourth-power algorithm of Exercise 12.17.
Exercise 12.19. This problem explores a receiver that requires carrier recovery,
timing recovery, and automatic gain correction in order to function properly.
Hint: you may want to peek ahead to Exercise 15.1.
a. Write a carrier-recovery routine for the signal generated by BigEx1.m. You
may want to use the receiver in BigIdeal.m for inspiration. Pay particular
attention to the given parameters (Hint: the sampling rate is below the IF
frequency, so the receiver front end eectively employs subsampling). Describe
the carrier-recovery method used and plot the tracking of . If the receiver
employs preprocessing, take care to design the BPF appropriately.

266

Chapter 12. Timing Recovery

b. Add a timing-recovery algorithm to the receiver so that the receiver samples

at the appropriate time. State the timing-recovery method used and plot the
tracking of .
c. Add an AGC to the code and decode the message. Does the code recover the
message appropriately? Calculate the error rate of the receiver.

12.5

Two Examples
This section presents two examples in which timing recovery plays a signicant
role. The rst looks at the behavior of the algorithms in the nonideal setting.
When there is channel ISI, the answer to which the algorithms converge is not
the same as in the ISI-free setting. This happens because the ISI of the channel
causes an eective delay in the energy that the algorithm measures. The second example shows how the timing-recovery algorithms can be used to estimate
(slow) changes in the optimal sampling time. When these changes occur linearly,
they are eectively a change in the underlying period, and the timing-recovery
algorithms can be used to estimate the oset of the period in the same way that
the phase estimates of the PLL in Section 10.6 can be used to nd a (small)
frequency oset in the carrier.

Example 12.3. Modify the simulation in clockrecDD.m by changing the channel:

chan =[1 0 . 7 0 0 . 5 ] ;

% T/m c h a n n e l

With an oversampling of m=2, 2-PAM constellation, and beta=0.5, the output

of the output-power-maximization algorithm clockrecOP.m is shown in Figure
12.12. With these parameters, the iteration begins in a closed-eye situation.
Because of the channel, no single timing parameter can hope to achieve a perfect
1 outcome. Nonetheless, by nding a good compromise position (in this case,
converging to an oset of about 0.6), the hard decisions are correct once the eye
has opened (which rst occurs around iteration 500).
Example 12.3 shows that the presence of ISI changes the convergent value of
the timing-recovery algorithm. Why is this?
Suppose rst that the channel was a pure delay. (For instance, set chan=[0
1] in Example 12.3). Then the timing algorithm will change the estimates tau
(in this case, by one) to maximize the output power to account for the added
delay. When the channel is more complicated, the timing recovery again moves
the estimates to that position which maximizes the output power, but the actual
value attained is a weighted version of all the taps. For example, with chan=[1
1], the energy is maximized halfway between the two taps and the answer is
oset by 0.5. Similarly, with chan=[3 1], the energy is located a quarter of the
way between the taps and the answer is oset by 0.25. In general, the oset is
(roughly) proportional to the size of the taps and their delay.

Offset estimates

Symbol estimates

Timing Recovery

267

Constellation diagram

3
2
1
0
1
2
3
0.8

Figure 12.12 Output of the

program clockrecOP.m,
modied for Example 12.3,
shows the constellation history
in the top plot. The trajectory
of the oset estimation is
shown in the bottom plot.

0.4
0
0

1000

2000

3000
Iterations

4000

5000

To see the general situation, consider the received analog signal due to a single
symbol triggering the pulse-shape lter and passing through a channel with ISI.
An adjustment in the baud-timing setting at the receiver will sample at slightly
dierent points on the received analog signal. A change in is eectively equivalent to a change in the channel ISI. This will be dealt with in Chapter 13 when
designing equalizers.

Example 12.4. With the signal generated as in clockrecDD.m on page 258, the
following code resamples (using sinc interpolation) the received signal to simulate
a change in the underlying period by a factor of fac.
Listing 12.5. clockrecperiod.m resample to change the period
f a c = 1 . 0 0 0 1 ; z=zeros ( s i z e ( x ) ) ;
t=l +1: f a c : length ( x)2 * l ;
f o r i=l +1: length ( t )
z ( i )= i n t e r p s i n c ( x , t ( i ) , l ) ;
end
x=z ;

%
%
%
%
%
%

p e r c e n t change i n p e r i o d
v e c t o r o f new t i m e s
r e s a m p l e x a t new r a t e
to c r e a t e r e c e i v e d s i g n a l
with p e r i o d o f f s e t
relabel signal

If this code is followed by one of the timing-recovery schemes, then the timing
parameter follows the changing period. For instance, in Figure 12.13, the timing
estimation converges rapidly to a line with slope that is proportional to the
dierence in period between the assumed value of the period at the receiver and
the actual value used at the transmitter.
Thus, the standard timing-recovery algorithms can handle the case in which
the clock periods at the transmitter and receiver are somewhat dierent. More
accurate estimates could be made using two timing-recovery algorithms analogous to the dual-carrier recovery structure of Section 10.6.2 or by mimicking
the second-order lter structure of the PLL in the article Analysis of the Phase
Locked Loop, which can be found on the website. There are also other common
timing-recovery algorithms such as the earlylate method, the method of Mueller
and M
uller, and band-edge timing algorithms.

Symbol estimates

Chapter 12. Timing Recovery

Offset estimates

268

Constellation history

1.5
1
0.5
0
0.5
1
1.5
1
0
1
2
3
4

Figure 12.13 Output of

clockrecperiod.m as modied for

Example 12.4 shows the
constellation history in the top
plot and the trajectory of the
oset estimation in the bottom.
The slope of the estimates is
proportional to the dierence
between the nominal and the
actual clock period.
0

0.5

1.5
Iterations

x104

Exercise 12.20. Modify clockrecOP.m to implement one of the alternative

performance functions of Figure 12.11: avg{|x[k]|}, avg{x2 [k]}, or avg{(x2 [k]
1)2 }.
Exercise 12.21. Modify clockrecOP.m by changing the channel as in Example 12.3. Use dierent values of beta in the SRRC pulse-shape routine. How does
this aect the convergence speed of the algorithm? Do dierent pulse shapes
change the convergent value?
Exercise 12.22. Investigate how the error surface depends on the input signal.
a. Draw the error surface for the output-energy-maximization timing-recovery
algorithm when the inputs are binary 1.
b. Draw the error surface when the inputs are drawn from the 4-PAM constellation, for the case in which the symbol 3 never occurs.
Exercise 12.23. Imitate Example 12.3 using a channel of your own choosing.
Do you expect that the eye will always be able to open?
Exercise 12.24. Instead of the ISI channel used in Example 12.3, include a
white-noise channel. How does this change the timing estimates?
Exercise 12.25. Explore the limits of the period tracking in Example 12.4.
How large can fac be made and still have the estimates converge to a line?
What happens to the cluster variance when the estimates cannot keep up? Does
it help to increase the size of the stepsize mu?

For Further Reading

A comprehensive collection of timing- and carrier-recovery schemes can be found
in the following two texts:

Timing Recovery

269

r H. Meyr, M. Moeneclaey, and S. A. Fechtel, Digital Communication Receivers,

Wiley, 1998;

r J. A. C. Bingham, The Theory and Practice of Modem Design, Wiley Interscience, 1988.

13 Linear Equalization

When all is well in the receiver, there is no interaction between successive symbols; each symbol arrives and is decoded independently of all others. But when
symbols interact, when the waveform of one symbol corrupts the value of a nearby
symbol, then the received signal becomes distorted. It is dicult to decipher
the message from such a received signal. This impairment is called intersymbol interference and was discussed in Chapter 11 in terms of non-Nyquist pulse
shapes overlapping in time. This chapter considers another source of interference
between symbols that is caused by multipath reections (or frequency-selective
dispersion) in the channel.
When there is no intersymbol interference (from a multipath channel, from
imperfect pulse shaping, or from imperfect timing), the impulse response of the
system from the source to the recovered message has a single nonzero term.
The amplitude of this single spike depends on the transmission losses, and the
delay is determined by the transmission time. When there is intersymbol interference caused by a multipath channel, this single spike is scattered, duplicated
once for each path in the channel. The number of nonzero terms in the impulse
response increases. The channel can be modeled as a nite-impulse-response,
linear lter C, and the delay spread is the total time interval during which reections with signicant energy arrive. The idea of the equalizer is to build (another)
lter in the receiver that counteracts the eect of the channel. In essence, the
equalizer must unscatter the impulse response. This can be stated as the goal
of designing the equalizer E so that the impulse response of the combined channel and equalizer CE has a single spike. This can be cast as an optimization
problem, and can be solved using techniques familiar from Chapters 6, 10, and
12.
The transmission path may also be corrupted by additive interferences such
as those caused by other users. These noise components are usually presumed
to be uncorrelated with the source sequence and they may be broadband or
narrowband, in band or out of band relative to the bandlimited spectrum of the
source signal. Like the multipath channel interference, they cannot be known to

Linear Equalization

271

Noise and interferers

Digital
source

Pulse
shaping

Analog
channel

Received
analog
signal

Sampled
received
signal
Ts

Linear
digital
equalizer

Decision
device

Figure 13.1 The baseband linear (digital) equalizer is intended to (automatically)

cancel out unwanted eects of the channel and to cancel out certain kinds of additive
interferences.

the system designer in advance. The second job of the equalizer is to reject such
additive narrowband interferers by designing appropriate linear notch lters onthe-y. At the same time, it is important that the equalizer does not unduly
enhance the broadband noise.
The signal path of a baseband digital communication system is shown in Figure 13.1, which emphasizes the role of the equalizer in trying to counteract the
eects of the multipath channel and the additive interference. As in previous
chapters, all of the inner parts of the system are assumed to operate precisely:
thus, the upconversion and downconversion, the timing recovery, and the carrier synchronization (all those parts of the receiver that are not shown in Figure
13.1) are assumed to be awless and unchanging. Modeling the channel as a timeinvariant FIR lter, the next section focuses on the task of selecting the coecients in the block labeled linear digital equalizer, with the goal of removing the
intersymbol interference and attenuating the additive interferences. These coecients are to be chosen using the sampled received signal sequence and (possibly)
knowledge of a prearranged training sequence. While the channel may actually
be time-varying, the variations are often much slower than the data rate, and
the channel can be viewed as (eectively) time-invariant over small time scales.
This chapter suggests several dierent ways in which the coecients of the
equalizer can be chosen. The rst procedure, in Section 13.2.1, minimizes the
square of the symbol-recovery error1 over a block of data, which can be done
using a matrix pseudoinversion. Minimizing the (square of the) error between the
received data values and the transmitted values can also be achieved using an
adaptive element, as detailed in Section 13.3. When there is no training sequence,
other performance functions are appropriate, and these lead to equalizers such as
the decision-directed approach in Section 13.4 and the dispersion-minimization
method in Section 13.5. The adaptive methods considered here are only modestly
complex to implement, and they can potentially track time variations in the
channel model, assuming that the changes are suciently slow.
1

This is the error between the equalizer output and the transmitted symbol, and is known
whenever there is a training sequence.

272

Chapter 13. Linear Equalization

13.1

Multipath Interference
The villains of this chapter are multipath and other additive interferers. Both
should be familiar from Section 4.1.
The distortion caused by an analog wireless channel can be thought of as a
combination of scaled and delayed reections of the original transmitted signal.
These reections occur when there are dierent paths from the transmitting
antenna to the receiving antenna. Between two microwave towers, for instance,
the paths may include one along the line of sight, reections from nearby hills,
and bounces from a eld or lake between the towers. For indoor digital TV
reception, there are many (local) time-varying reectors, including people in the
receiving room, and nearby vehicles. The strength of the reections depends on
the physical properties of the reecting objects, while the delay of the reections
is primarily determined by the length of the transmission path. Let u(t) be the
transmitted signal. If N delays are represented by 1 , 2 , . . . , N , and the
strength of the reections is 1 , 2 , . . . , N , then the received signal is
y(t) = 1 u(t 1 ) + 2 u(t 2 ) + + N u(t N ) + (t),

(13.1)

where (t) represents additive interferences. This model of the transmission channel has the form of a nite-impulse-response lter, and the total length of time
N 1 over which the impulse response is nonzero is called the delay spread
of the physical medium.
This transmission channel is typically modeled digitally assuming a xed sampling period Ts . Thus, (13.1) is approximated by
y(kTs ) = a0 u(kTs ) + a1 u((k 1)Ts ) + + an u((k n)Ts ) + (kTs ). (13.2)
In order for the model (13.2) to closely represent the system (13.1), the total time
over which the impulse response is nonzero (the time nTs ) must be at least as
large as the maximum delay N . Since the delay is not a function of the symbol
period Ts , smaller values of Ts require more terms in the lter (i.e., larger n).
For example, consider a sampling interval of Ts 40 ns (i.e., a transmission
rate of 25 MHz). A delay spread of approximately 4 s would correspond to 100
taps in the model (13.2). Thus, at any time instant, the received signal would
be a combination of (up to) 100 data values. If Ts were increased to 0.4 s
(i.e., 2.5 MHz), only 10 terms would be needed, and there would be interference
with only the 10 nearest data values. If Ts were larger than 4 s (i.e., 0.25
MHz), only one term would be needed in the discrete-time impulse response.
In this case, adjacent sampled symbols would not interfere. Such nite-duration
impulse-response models as (13.2) can also be used to represent the frequencyselective dynamics that occur in the wired local end-loop in telephony, and other
(approximately) linear, nite-delay-spread channels.
The design objective of the equalizer is to undo the eects of the channel and to
remove the interference. Conceptually, the equalizer attempts to build a system
that is a delayed inverse of (13.2), removing the intersymbol interference while

Linear Equalization

273

simultaneously rejecting additive interferers uncorrelated with the source. If the

interference (kTs ) is unstructured (for instance white noise) then there is little
that a linear equalizer can do to remove it. But when the interference is highly
structured (such as narrowband interference from another user), then the linear
lter can often notch out the oending frequencies.
As shown in Example 12.3 of Section 12.5, the solution for the optimal sampling times found by the clock-recovery algorithms depends on the ISI in the
channel. Consequently, the digital model (such as (13.2)) formed by sampling an
analog transmission path (such as (13.1)) depends on when the samples are taken
within each period Ts . To see how this can happen in a simple case, consider a
two-path transmission channel
(t) + 0.6(t ),
where is some fraction of Ts . For each transmitted symbol, the received signal
will contain two copies of the pulse shape p(t), the rst undelayed and the second
delayed by and attenuated by a factor of 0.6. Thus, the receiver sees
c(t) = p(t) + 0.6p(t ).
This is shown in Figure 13.2 for = 0.7Ts . The clock-recovery algorithms cannot
separate the individual copies of the pulse shapes. Rather, they react to the
complete received shape, which is their sum. The power maximization will locate
the sampling times at the peak of this curve, and the lattice of sampling times
will be dierent from what would be expected without ISI. The eective (digital)
channel model is thus a sampled version of c(t). This is depicted in Figure 13.2
by the small circles that occur at Ts -spaced intervals.
In general, an accurate digital model for a channel depends on many things:
the underlying analog channel, the pulse shaping used, and the timing of the
sampling process. At rst glance, this seems like it might make designing an
equalizer for such a channel almost impossible. But there is good news. No
matter what timing instants are chosen, no matter what pulse shape is used, and
no matter what the underlying analog channel may be (as long as it is linear),
there is an FIR linear representation of the form (13.2) that closely models its
behavior. The details may change, but it is always a sampling of the smooth curve
(like c(t) in Figure 13.2) that denes the digital model of the channel. As long
as the digital model of this channel does not have deep nulls (i.e., a frequency
response that practically zeroes out some important band of frequencies), there
is a good chance that the equalizer can undo the eects of the channel.

13.2

Trained Least-Squares Linear Equalization

When there is a training sequence available (for instance, in the known frame
information that is used in synchronization), this can also be used to help build
or train an equalizer. The basic strategy is to nd a suitable function of the

274

Chapter 13. Linear Equalization

Lattice of Ts-spaced
optimal sampling times
with ISI
Lattice of Ts-spaced optimal
sampling times with no ISI
Sum of received pulses
c(t) = p(t) + 0.6 p(t )

p(t)
0.6 p(t )

0
The digital channel
model is given by
Ts-spaced samples of c(t)
Figure 13.2 The optimum sampling times (as found by the energy-maximization

algorithm) dier when there is ISI in the transmission path, and change the eective
digital model of the channel.
Additive
interferers
Source
s[k]
Channel

Received
signal r[k]

Equalizer
output y[k]
Equalizer

Impulse response f
Training signal
Delay

Error
e[k]

Figure 13.3 The problem of linear equalization is to nd a linear system f that undoes
the eects of the channel while minimizing the eects of the interferences.

unknown equalizer parameters that can be used to dene an optimization problem. Then, applying the techniques of Chapters 6, 10, and 12, the optimization
problem can be solved in a variety of ways.

13.2.1

A Matrix Description
The linear equalization problem is depicted in Figure 13.3. A prearranged training sequence s[k] is assumed known at the receiver. The goal is to nd an FIR
lter (called the equalizer) so that the output of the equalizer is approximately
equal to the known source, though possibly delayed in time. Thus, the goal is to
choose the impulse response fi so that y[k] s[k ] for some specic .

Linear Equalization

r[k]

z1
f0

z1

...

z1

f1

275

fn

...

y[k]

Figure 13.4 The direct-form FIR lter of Equation (13.3) can be pictured as a tapped
delay line where each z 1 block represents a time delay of one symbol period. The
impulse response of the lter is f0 , f1 , . . ., fn .

The inputoutput behavior of the FIR linear equalizer can be described as the
convolution
n

y[k] =
fj r[k j],
(13.3)
j=0

where the lower index on j can be no lower than zero (or else the equalizer
is noncausal; that is, it can illogically respond to an input before the input is
applied). This convolution is illustrated in Figure 13.4 as a direct-form FIR or
tapped delay line.
The summation in (13.3) can also be written (e.g., for k = n + 1) as the inner
product of two vectors

f0
f1

(13.4)
y[n + 1] = [r[n + 1], r[n], . . . , r[1]] . .
..
fn
Note that y[n + 1] is the earliest output that can be formed given no knowledge
of r[i] for i < 1. Incrementing the time index in (13.4) gives

f0
f1

y[n + 2] = [r[n + 2], r[n + 1], . . . , r[2]] .
..
fn
and

f0
f1

y[n + 3] = [r[n + 3], r[n + 2], . . . , r[3]] . .
..
fn

Observe that each of these uses the same equalizer parameter vector. Concatenating p n of these measurements into one matrix equation over the available

276

Chapter 13. Linear Equalization

data set for i = 1 to p gives

r[n + 1] r[n]
y[n + 1]
y[n + 2] r[n + 2] r[n + 1]

y[n + 3] r[n + 3] r[n + 2]

=

..
..
..

.
.
.
r[p]
r[p 1]
y[p]

r[1]
f0
r[2]
f1

r[3]
. ,

.
..
.
.
fn
r[p n]

(13.5)

or, with the appropriate matrix denitions,

Y = RF.

(13.6)

Note that R has a special structure, namely that the entries along each diagonal
are the same. R is known as a Toeplitz matrix and the toeplitz command in
Matlab makes it easy to build matrices with this structure.

13.2.2

Source Recovery Error

The delayed source recovery error is
e[k] = s[k ] y[k]

(13.7)

for a particular . This section shows how the source recovery error can be used
to dene a performance function that depends on the unknown parameters fi .
Calculating the parameters that minimize this performance function provides a
good solution to the equalization problem.
Dene

e[n + 1]
s[n + 1 ]
e[n + 2]
s[n + 2 ]

s[n + 3 ]
(13.8)
S=
and E = e[n + 3] .

..
..

.
.
e[p]
s[p ]
Using (13.6), write
E = S Y = S RF.

(13.9)

As a measure of the performance of the fi in F , consider

JLS =

p


e2 [i].

(13.10)

i=n+1

JLS is nonnegative since it is a sum of squares. Minimizing such a summed

squared delayed source recovery error is a common objective in equalizer design,
since the fi that minimize JLS cause the output of the equalizer to become close
to the values of the (delayed) source.

Linear Equalization

277

Given (13.8) and (13.9), JLS in (13.10) can be written as

JLS = E T E = (S RF )T (S RF )
= S T S (RF )T S S T RF + (RF )T RF.

(13.11)

Because JLS is a scalar, (RF )T S and S T RF are also scalars. Since the transpose
of a scalar is equal to itself, (RF )T S = S T RF , and (13.11) can be rewritten as
JLS = S T S 2S T RF + (RF )T RF.

(13.12)

The issue is now one of choosing the n + 1 entries of F to make JLS as small as
possible.

13.2.3

The Least-Squares Solution

Dene the matrix
= [F (RT R)1 RT S]T (RT R)[F (RT R)1 RT S]
= F T (RT R)F S T RF F T RT S + S T R(RT R)1 RT S.
The purpose of this denition is to rewrite (13.12) in terms of :
JLS = + S T S S T R(RT R)1 RT S
= + S T [I R(RT R)1 RT ]S.

(13.13)

Since S T [I R(RT R)1 RT ]S is not a function of F , the minimum of JLS

occurs at the F that minimizes . This occurs when
F = (RT R)1 RT S,

(13.14)

assuming that (RT R)1 exists.2 The corresponding minimum achievable by JLS
at F = F is the summed squared delayed source recovery error. This is the
remaining term in (13.13); that is,
min
JLS
= S T [I R(RT R)1 RT ]S.

(13.15)

The formulas for the optimum F in (13.14) and the associated minimum
achievable JLS in (13.15) are for a specic . To complete the design task, it
is also necessary to nd the optimal delay . The most straightforward approach
is to set up a series of S = RF calculations, one for each possible , to compute
min
the associated values of JLS
, and pick the delay associated with the smallest
one.
This procedure is straightforward to implement in Matlab, and the program
LSequalizer.m allows you to play with the various parameters to get a feel for
their eect. Much of this program will be familiar from openclosed.m. The rst
2

A matrix is invertible as long as it has no eigenvalues equal to zero. Since RT R is a quadratic

form it has no negative eigenvalues. Thus, all eigenvalues must be positive in order for it to
be invertible.

278

Chapter 13. Linear Equalization

three lines dene a channel, create a binary source, and then transmit the source
through the channel using the filter command. At the receiver, the data are
put through a quantizer, and then the error is calculated for a range of delays.
The new part is in the middle.
Listing 13.1. LSequalizer.m nd a LS equalizer f for the channel b
b=[0.5 1 0.6];
% def i ne channel
m=1000; s=sign ( randn ( 1 ,m) ) ;
% binary source of length m
r=f i l t e r ( b , 1 , s ) ;
% output o f channel
n=3;
% length of equal i zer 1
d e l t a =3;
% u s e d e l a y <=n * l e n g t h ( b )
p=length ( r ) d e l t a ;
R=t o e p l i t z ( r ( n+1:p ) , r ( n + 1 : 1 : 1 ) ) ;
% b u i l d m a tr i x R
S=s ( n+1d e l t a : pd e l t a ) ;
% and v e c t o r S
f=inv (R * R) * R * S
% calculate equalizer f
Jmin=S * SS * R* inv (R * R) * R * S
% Jmin f o r t h i s f and d e l t a
y=f i l t e r ( f , 1 , r ) ;
% equalizer is a f i l t e r
dec=sign ( y ) ;
% q u a n t i z e and f i n d e r r o r s
e r r =0.5 *sum( abs ( dec ( d e l t a +1:m) s ( 1 :md e l t a ) ) )

The variable n denes the length of the equalizer, and delta denes the delay
that will be used in constructing the vector S dened in (13.8) (observe that
delta must be positive and less than or equal to n). The Toeplitz matrix R
is dened in (13.5) and (13.6), and the equalizer coecients f are computed
as in (13.14). The value of minimum achievable performance is Jmin, which is
calculated as in (13.15). To demonstrate the eect of the equalizer, the received
signal r is ltered by the equalizer coecients, and the output is then quantized.
If the equalizer has done its job (i.e., if the eye is open), then there should be
some shift sh at which no errors occur.
For example, using the default channel b = [0.5, 1, 0.6], and length-4 equalizer
(n=3), four values of the delay delta give
equalizer f
delay delta Jmin
0
832 {0.33, 0.027, 0.070, 0.01}
134 {0.66, 0.36, 0.16, 0.08}
1
2
30 {0.28, 0.65, 0.30, 0.14}
45 {0.1, 0.27, 0.64, 0.3}
3

(13.16)

The best equalizer is the one corresponding to a delay of 2, since this Jmin is the
smallest. In this case, however, any of the last three will open the eye. Observe
that the number of errors (as reported in err) is zero when the eye is open.
Exercise 13.1. Plot the frequency response (using freqz) of the channel b
in LSequalizer.m. Plot the frequency response of each of the four equalizers
found by the program. For each channel/equalizer pair, form the product of the
magnitude of the frequency responses. How close are these products to unity?
Exercise 13.2. Add (uncorrelated, normally distributed) noise into the simulation using the command r=filter(b,1,s)+sd*randn(size(s)).

Linear Equalization

279

a. For the equalizer with delay 2, what is the largest sd you can add and still
have no errors?
b. Make a plot of Jmin as a function of sd.
c. Now try the equalizer with delay 1. What is the largest sd you can add, and
still have no errors?
d. Which is a better equalizer?
Exercise 13.3. Use LSequalizer.m to nd an equalizer that can open the eye
for the channel b = [1, 1, 0.8, 0.3, 1, 1].
a.
b.
c.
d.
e.
f.

What equalizer length n is needed?

What delays delta give zero error at the output of the quantizer?
What is the corresponding Jmin?
Plot the frequency response of this channel.
Plot the frequency response of your equalizer.
Calculate and plot the product of the two.

Exercise 13.4. Modify LSequalizer.m to generate a source sequence from the

alphabet 1, 3. For the default channel [0.5, 1, 0.6], nd an equalizer that
opens the eye.
a.
b.
c.
d.

What equalizer length n is needed?

What delays delta give zero error at the output of the quantizer?
What is the corresponding Jmin?
Is this a fundamentally easier or more dicult task than when equalizing a
binary source?
e. Plot the frequency response of the channel and of the equalizer.

There is a way to convert the exhaustive search over all the delays in the
previous approach into a single matrix operation. Construct the (p ) ( +
1) matrix of training data

s[ + 1] s[]
s[1]
s[ + 2] s[ + 1]
s[2]

(13.17)
S =
,
..
..
..

.
.
.
s[p]

s[p 1] s[p ]

where species the number of delays that will be searched, from = 0 to

= . The (p ) (n + 1) matrix of received data is

r[ + 1] r[]
r[ n + 1]
r[ + 2] r[ + 1] r[ n + 2]

=
R
(13.18)

,
..
..
..

.
.
.
r[p]

r[p 1]

r[p n]

where each column corresponds to one of the possible delays. Note that > n is
required in order to keep the lowest index of r[] positive. In the (n + 1) ( + 1)

280

Chapter 13. Linear Equalization

matrix

F =

f00
f10
..
.
fn0

f01 f0
f11 f1
..
..
.
.
fn1 fn

each column is a set of equalizer parameters, one corresponding to each of the

to nd F . The column of F that
possible delays. The strategy is to use S and R
results in the smallest value of the performance function JLS is then the optimal
receiver at the optimal delay.
The jth column of F corresponds to the equalizer parameter vector choice
with this jth column from F is intended to
for = j 1. The product of R
The least-squares solution of S R
F is
approximate the jth column of S.
T R)
1 R
T S,

F = (R

(13.19)

(i.e., n + 1) must be less than or equal to

where the number of columns of R
(i.e., p ) for (R
T R)
1 to exist. Consequently, p
the number of rows of R
n + 1 implies that p > n + . If so, the minimum value associated with a
particular column of F (e.g., F ) is, from (13.15),
min,
1 R
T ]S ,
R
T R)
JLS
= ST [I R(

Thus, these
where S is the th column of S.

min,
JLS

(13.20)

are all along the diagonal of

R
T R)

1 R
T ]S.
= ST [I R(

(13.21)

Hence, the minimum value on the diagonal of (e.g., at the (j, j)th entry)
corresponds to the optimum .

Example 13.1. Consider the low-order example with n = 1 (so F has two parameters), = 2 (so > n), and p = 5 (so p > n + ). Thus,



s[3] s[2] s[1]
r[3] r[2]
= r[4] r[3] , and F = f00 f01 f02 .
S = s[4] s[3] s[2] , R
f10 f11 f12
s[5] s[4] s[3]
r[5] r[4]
For the example, assume that the true channel is
r[k] = ar[k 1] + bs[k 1].
A two-tap equalizer F = [f0 , f1 ]T can provide perfect equalization for = 1 with
f0 = 1/b and f1 = a/b, since
y[k] = f0 r[k] + f1 r[k 1] =
=

1
[r[k] ar[k 1]]
b

1
[ar[k 1] + bs[k 1] ar[k 1]] = s[k 1].
b

Linear Equalization

281

Consider
{s[1], s[2], s[3], [s4], s[5]} = {1, 1, 1, 1, 1},
which results in

1 1 1
S = 1 1 1 .
1 1 1

With a = 0.6, b = 1, and r[1] = 0.8. Similarly,

r[2] = ar[1] + bs[1] = 0.48 + 1 = 1.48,
r[3] = 0.888 1 = 0.112,
r[4] = 1.0672,
r[5] = 0.3597.

Example 13.2. The eect of channel noise can be simulated by rounding these
values for r in composing

0.1 1.5
= 1.1 0.1 .
R
0.4 1.1
Thus, from (13.21),

1.2848 0.0425 0.4778

= 0.0425 0.0014 0.0158 ,
0.4778 0.0158 0.1777
and from (13.19),
F =

1.1184 0.9548 0.7411

0.2988 0.5884 0.8806

Since the second diagonal term in is the smallest diagonal term, = 1 is the
optimum setting (as expected) and the second column of F is the minimum
summed squared delayed recovery error solution (i.e., f0 = 0.9548 ( 1/b = 1)
and f1 = 0.5884 ( a/b = 0.6)).
With a better received signal measurement, for instance,

0.11
1.48
= 1.07 0.11 ,
R
0.36 1.07
the diagonal of is [1.3572, 0.0000, 0.1657] and the optimum delay is again
= 1, and the optimum equalizer settings are 0.9960 and 0.6009, which give a
(measured without
better t to the ideal noise-free answer. Innite precision in R

282

Chapter 13. Linear Equalization

channel noise or other interferers) produces a perfect t to the true f0 and f1

and a zeroed delayed source recovery error.

13.2.4

Summary of Least-Squares Equalizer Design

The steps of the linear FIR equalizer design strategy are as follows.
1.
2.
3.
4.
5.
6.
7.

Select the order n for the FIR equalizer in (13.3).

Select the maximum of candidate delays > n used in (13.17) and (13.18).
Utilize a set of p training signal samples {s[1], s[2], . . . , s[p]} with p > n + .
Obtain a corresponding set of p received signal samples {r[1], r[2], . . . , r[p]}.
Compose S in (13.17).
in (13.18).
Compose R
has poor conditioning induced by any (near) zero eigenTR
Check whether R
values. Matlab will return a warning (or an error) if the matrix is too close
to singular.3
8. Compute F from (13.19).
9. Compute by substituting F into (13.21), rewritten as
F ].
= ST [S R
10. Find the minimum value on the diagonal of . This index is + 1. The associated diagonal element of is the minimum achievable summed squared

delayed source recovery error pi=+1 e2 [i] over the available data record.
11. Extract the ( + 1)th column of the previously computed F . This is the
impulse response of the optimum equalizer.
12. Test the design. Test it on synthetic data, and then on measured data (if
available). If it is inadequate, repeat the design stage, perhaps increasing n
or twiddling some other designer-selected quantity.
This procedure, along with three others that will be discussed in the ensuing
sections, is available on the website in the program dae.m. Combining the various approaches makes it easier to compare their behaviors in the examples of
Section 13.6.

13.2.5

Complex Signals and Parameters

The preceding development assumes that the source signal and channel, and
therefore the received signal, equalizer, and equalizer output, are all real-valued.
However, the source signal and channel may be modeled as complex-valued when
using modulations such as QAM of Section 5.3. This is explored in some detail
3

The condition number (i.e., the maximum eigenvalue divided by the minimum eigenvalue) of
should be checked. If the condition number is extremely large, start again with dierent
T R
R
then most likely the channel
T R,
{s[]}. If all choices of {s[]} result in poorly conditioned R
has deep nulls that prohibit the successful application of a T -spaced linear equalizer.

Linear Equalization

283

in the document A Digital Quadrature Amplitude Modulation Radio, which can

be found on the website. The same basic strategy for equalizer design can also
be used in the complex case.
Consider a complex delayed source recovery error

where j =

e[k] = eR [k] + jeI [k],

1. Consider its square,
e2 [k] = e2R [k] + 2jeR [k]eI [k] e2I [k],

which is typically complex-valued, and potentially real-valued and negative when

eR 0. Thus, a sum of e2 is no longer a suitable measure of performance since
|e| might be nonzero but its squared average might be zero.
Instead, consider the product of a complex e with its complex conjugate e =
eR jeI ; that is,
e[k]e [k] = e2R [k] jeR [k]eI [k] + jeR [k]eI [k] j 2 e2I [k] = e2R [k] + e2I [k].
In vector form, the summed squared error of interest is E H E (rather than
the E H E of (13.11)), where the superscript H denotes the operations of both
transposition and complex conjugation. Thus, (13.14) becomes
F = (RH R)1 RH S.
Note, that in implementing this renement in the Matlab code, the symbol pair
. implements a transpose, while alone implements a conjugate transpose.

13.2.6

Fractionally Spaced Equalization

The preceding development assumes that the sampled input to the equalizer
is symbol spaced with the sampling interval equal to the symbol interval of T
seconds. Thus, the unit delay in realizing the tapped-delay-line equalizer is T
seconds. Sometimes, the input to the equalizer is oversampled such that the
sample interval is shorter than the symbol interval, and the resulting equalizer
is said to be fractionally spaced. The same kinds of algorithms and solutions can
be used to calculate the coecients in fractionally spaced equalizers as are used
for T -spaced equalizers. Of course, details of the construction of the matrices
will necessarily dier due to the structural dierences.
corresponding to S and R
The more rapid sampling allows greater latitude in the ordering of the blocks in
the receiver. This is discussed at length in Equalization on the website.
Exercise 13.5. Consider the multiuser system shown in Figure 13.5. Both users
transmit binary 1 PAM signals that are independent and equally probable symbols. The signal from the rst user is distorted by a frequency-selective channel
with impulse response
h1 [k] = [k] + b[k 1].

284

Chapter 13. Linear Equalization

User 1
s1[k] Channel 2
h1[k]
User 2
s2[k] Channel 1
h2[k]

Figure 13.5 Baseband

r[k] Equalizer x[k] Decision

f [k]
device

y[k]

communication system with

multiuser interference used in
Exercise 13.5.

The signal from the second (interfering) user is scaled by h2 [k] = c[k]. The
dierence equation generating the received signal r is
r[k] = s1 [k] + bs1 [k 1] + cs2 [k].
The dierence equation describing the equalizer inputoutput behavior is
x[k] = r[k] + dr[k 1].
The decision device is the sign operator.
a. Write out the dierence equation relating s1 [k] and s2 [k] to x[k].
b. Consider the performance function
JMSE =

1
N

k
0 +N

(s1 [k] x[k])2 .

k=k0 +1

For uncorrelated source symbol sequences, JMSE is minimized by minimizing

 2
i . For b = 0.8 and c = 0.3 compute the equalizer setting d that minimizes
JMSE .
c. For b = 0.8, c = 0.3, and d = 0 (i.e., equalizer-less reception), does the system
exhibit decision errors?
d. For b = 0.8, c = 0.3, and your d from part (b), does the system exhibit decision
errors? Does this system suer more severe decision errors than the equalizerless system of part (c)?

13.3

An Adaptive Approach to Trained Equalization

The block-oriented design of the previous section requires substantial computation even when the system delay is known since it requires calculating the
inverse of an (n + 1) (n + 1) matrix, where n is the largest delay in the FIR
linear equalizer. This section considers using an adaptive element to minimize
the average of the squared error,
JLMS =

1
avg{e2 [k]}.
2

(13.22)

Linear Equalization

285

Observe that JLMS is a function of all the equalizer coecients fi , since

e[k] = s[k ] y[k] = s[k ]

n


fj r[k j],

(13.23)

j=0

which combines (13.7) with (13.3), and where r[k] is the received signal at baseband after sampling. An algorithm for the minimization of JLMS with respect
to the ith equalizer coecient fi is

JLMS 
.
(13.24)
fi [k + 1] = fi [k]
fi fi =fi [k]
To create an algorithm that can easily be implemented, it is necessary to evaluate
this derivative with respect to the parameter of interest. This is


1 2
e [k]
avg{ 12 e2 [k]}
JLMS
e[k]
2
= avg e[k]
,
(13.25)
=
avg
fi
fi
fi
fi
where the approximation follows from (G.12) and the nal equality from the
chain rule (A.59). Using (13.23), the derivative of the source recovery error e[k]
with respect to the ith equalizer parameter fi is
e[k]
s[k ]  fj r[k j]
=

= r[k i],
fi
fi
fi
j=0
n

(13.26)

since s[k ]/fi = 0 and fj r[k j]/fi = 0 for all i = j. On substituting

(13.26) into (13.25) and then into (13.24), the update for the adaptive element
is
fi [k + 1] = fi [k] + avg{e[k]r[k i]}.
Typically, the averaging operation is suppressed since the iteration with small
stepsize itself has a lowpass (averaging) behavior. The result is commonly
called the least mean squares (LMS) algorithm for direct linear equalizer impulse
response coecient adaptation:
fi [k + 1] = fi [k] + e[k]r[k i].

(13.27)

This adaptive equalization scheme is illustrated in Figure 13.6.

When all goes well, the recursive algorithm (13.27) converges to the vicinity
of the block least-squares answer for the particular used in forming the delayed
recovery error. As long as is nonzero, if the underlying composition of the
received signal changes so that the error increases and the desired equalizer
changes, then the fi react accordingly. It is this tracking ability that earns it the
label adaptive.4
The following Matlab code implements an adaptive equalizer design. The
beginning and ending of the program are familiar from openclosed.m and
4

To provide tracking capability, the matrix solution of Section 13.2.1 could be recomputed for
successive data blocks, but this requires signicantly more computation.

286

Chapter 13. Linear Equalization

Sampled
received
signal r[k]

f [k]

Sign[]
y[k]

Equalizer

Adaptive
algorithm

e[k]

Decision
device

Performance
evaluation

s[k]
training
signal

Figure 13.6 A trained adaptive linear equalizer uses the dierence between the

received signal and a prespecied training sequence to drive the adaptation of the
coecients of the equalizer.

LSequalizer.m. The heart of the recursion lies in the for loop. For each new
data point, a vector is built containing the new value and the past n values of the
received signal. This is multiplied by f to make a prediction of the next source
symbol, and the error is the dierence between the prediction and the reality.
(This is the calculation of e[k] from (13.23).) The equalizer coecients f are then
updated as in (13.27).
Listing 13.2. LMSequalizer.m nd a LMS equalizer f for the channel b
b=[0.5 1 0.6];
m=1000; s=pam (m, 2 , 1 ) ;
r=f i l t e r ( b , 1 , s ) ;
n=4; f=zeros ( n , 1 ) ;
mu= . 0 1 ; d e l t a =2;
f o r i=n+1:m
r r=r ( i : 1 : i n + 1 ) ;
e=s ( i d e l t a ) r r * f ;
f=f+mu* e * r r ;
end

%
%
%
%
%
%
%
%
%

def i ne channel
binary source of length m
output o f channel
i n i t i a l i z e e q u a l i z e r at 0
s t e p s i z e and d e l a y d e l t a
iterate
vector of received s i gnal
calculate error
update e q u a l i z e r c o e f f i c i e n t s

As with the matrix approach, the default channel b=[0.5 1 -0.6] can be
equalized easily with a short equalizer (one with a small n). The convergent values of the f are very close to the nal values of the matrix approach; that is, for a
given channel, the value of f given by LMSequalizer.m is very close to the value
found using LSequalizer.m. A design consideration in the adaptive approach to
equalization involves the selection of the stepsize. Smaller stepsizes mean that
the trajectory of the estimates is smoother (tends to reject noises better) but
it also results in a slower convergence and slower tracking when the underlying
solution is time-varying. Similarly, if the explicit averaging operation is retained,
longer averages imply smoother estimates but slower convergence. Similar tradeos appear in the block approach in the choice of block size: larger blocks average
the noise better, but give no details about changes in the underlying solution
within the time span covered by a block.
The following code EqualizerTest.m shows one way to verify that the equalizer has (or has not) done a good job. The code generates a new received signal
and passes it through the same channel and the nal converged equalizer from
LMSequalizer.m. The decision sequence dec (the sign of the data for a binary

Linear Equalization

287

transmission) is then compared with the original transmitted data s. Since the
delay of the equalizer cannot be known beforehand, the loop on sh tests all the
possible delays (shifts) that might occur. If one of these is zero (or very small)
then the equalizer is doing a good job.
Listing 13.3. EqualizerTest.m verify the operation of an equalizer f for the channel b
% F i r s t run L M S e q u a l i z e r .m t o s e t c h a n n e l b and e q u a l i z e r f
f i n a l e q=f ;
% test final f i l t e r f
m=1000;
% new data p o i n t s
s=pam (m, 2 , 1 ) ;
% new b i n a r y s o u r c e o f l e n g t h m
r=f i l t e r ( b , 1 , s ) ;
% output o f channel
y t=f i l t e r ( f , 1 , r ) ;
% u s e f i n a l f i l t e r f to t e s t
dec=sign ( r e a l ( y t ) ) ;
% quantization
f o r sh =0:n
% i f e q u a l i z e r i s working , one
e r r ( sh +1)=0.5 *sum( abs ( dec ( sh +1:m) s ( 1 :msh ) ) ) ;
end
% o f t h e s e d e l a y s has z e r o e r r o r

This trained adaptive approach, along with several others, is implemented in

the program dae.m, which is available on the website. Simulated examples of
LMS with training and other adaptive equalization methods are presented in
Section 13.6.
Exercise 13.6. Verify that, by proper choice of n and delta, the convergent
values of f in LMSequalizer.m are close to the values shown in (13.16).
Exercise 13.7. What happens in LMSequalizer.m when the stepsize parameter
mu is too large? What happens when it is too small?
Exercise 13.8. Add (uncorrelated, normally distributed) noise into the simulation using the command r=filter(b,1,s)+sd*randn(size(s)).
a. For the equalizer with delay 2, what is the largest sd you can add, and still
have no errors? How does this compare with the result from Exercise 13.2?
Hint: it may be necessary to simulate for more than the default m data points.
b. Now try the equalizer with delay 1. What is the largest sd you can add, and
still have no errors?
c. Which is a better equalizer?
Exercise 13.9. Use LMSequalizer.m to nd an equalizer that can open the eye
for the channel b = [1, 1, 0.8, 0.3, 1, 1].
a. What equalizer length n is needed?
b. What delays delta give zero error in the output of the quantizer?
c. How does the answer compare with the design in Exercise 13.3?
Exercise 13.10. Modify LMSequalizer.m and EqualizerTest.m to generate a
source sequence from the alphabet 1, 3. For the default channel [0.5, 1, 0.6],
nd an equalizer that opens the eye.

288

Chapter 13. Linear Equalization

a. What equalizer length n is needed?

b. What delays delta give zero error in the output of the quantizer?
c. Is this a fundamentally easier or more dicult task than when equalizing a
binary source?
d. How does the answer compare with the design in Exercise 13.4?
Exercise 13.11. The trained adaptive equalizer updates its impulse response
coecients fi [k] via (13.27) using a gradient descent of the performance function
JLMS (13.22). With channels that have deep nulls in their frequency response
and a high SNR in the received signal, the minimization of (the average of)
JLMS results in a large spike in the equalizer frequency response (in order for
their product to be unity at the frequency of the channel null). Such large spikes
in the equalizer frequency response also amplify channel noise. To inhibit the
resulting large values of fi needed to create a frequency response with segments
of high gain, consider a cost function that also penalizes the sum of the squares
of the fi , e.g.,
&
'
N
1

1
2
2
e [k] +
fi [k]
2
i=0
Derive the associated adaptive-element update law corresponding to this performance function.

13.4

Decision-Directed Linear Equalization

During the training period, the communication system does not transmit any
message data. Commonly, a block of training data is followed by a block of
message data. The fraction of time devoted to training should be small, but can
be up to 20% in practice. If it were possible to adapt the equalizer parameters
without using the training data, the message-bearing (and revenue-generating)
capacity of the channel would be enhanced.
Consider the situation in which some procedure has produced an equalizer
setting that opens the eye of the channel. Thus, all decisions are perfect, but
the equalizer parameters might not yet be at their optimal values. In such a
case, the output of the decision device is an exact replica of the delayed source
(i.e., it is as good as a training signal). For a binary 1 source and decision
device that is a sign operator, the delayed source recovery error can be computed
as sign{y[k]} y[k], where y[k] is the equalizer output and sign{y[k]} equals
s[k ]. Thus, the trained adaptive equalizer of Figure 13.6 can be replaced by
the decision-directed equalizer shown in Figure 13.7. This converts (13.27) into
decision-directed LMS, which has the update
fi [k + 1] = fi [k] + (sign(y[k]) y[k])r[k i].

(13.28)

Linear Equalization

Sampled
received
signal r[k]

f [k]

Sign[]
y[k]

Equalizer

Adaptive
algorithm

289

e[k]

Decision
device

Performance
evaluation

Figure 13.7 A decision-directed adaptive linear equalizer uses the dierence between
the received signal and the output of the decision device to drive the adaptation of
the coecients of the equalizer.

When the signal s[k] is multilevel instead of binary, the sign function in (13.28)
can be replaced with a quantizer.
Exercise 13.12. Show that the decision-directed LMS algorithm (13.28)
can be derived as an adaptive element with performance function
(1/2) avg{(sign(y[k]) y[k])2 }. Hint: suppose that the derivative of the sign function dsign(x)/dx is zero everywhere.
Observe that the source signal s[k] does not appear in (13.28). Thus, no training signal is required for its implementation and the decision-directed LMS equalizer adaptation law of (13.28) is called a blind equalizer. Given its genesis, one
should expect decision-directed LMS to exhibit poor behavior when the assumption regarding perfect decisions is violated. The basic rule of thumb is that 5%
(or so) decision errors can be tolerated before decision-directed LMS fails to
converge properly.
The Matlab program DDequalizer.m has a familiar structure. The only code
changed from LMSequalizer.m is the calculation of the error term, which implements e[k] = sign{y[k]} y[k] rather than the LMS error (13.23), and the initialization of the equalizer. Because the equalizer must begin with an open eye,
f=0 is a poor choice. The initialization that follows starts all taps at zero except
for one in the middle that begins at unity. This is called the center-spike initialization. If the channel eye is open, then the combination of the channel and
equalizer will also have an open eye when initialized with the center spike. The
exercises ask you to explore the issue of nding good initial values for the equalizer parameters. As with the LMS equalizer, the code in EqualizerTest.m can
be used to test the operation of the converged equalizer.
Listing 13.4. DDequalizer.m nd a DD equalizer f for the channel b
b=[0.5 1 0.6];
m=1000; s=sign ( randn ( 1 ,m) ) ;
r=f i l t e r ( b , 1 , s ) ;
n=4; f =[0 1 0 0 ] ;
mu= . 1 ;
f o r i=n+1:m
r r=r ( i : 1 : i n + 1 ) ;

%
%
%
%
%
%
%

def i ne channel
binary source of length m
output o f channel
i ni t i a l i z e equalizer
stepsize
iterate
vector of received s i gnal

290

Chapter 13. Linear Equalization

e=sign ( f * r r ) f * r r ;
f=f+mu* e * r r ;
end

% calculate error
% update e q u a l i z e r c o e f f i c i e n t s

Exercise 13.13. Try the initialization f=[0 0 0 0] in DDequalizer.m. With

this initialization, can the algorithm open the eye? Try increasing m. Try changing
the stepsize mu. What other initializations will work?
Exercise 13.14. What happens in DDequalizer.m when the stepsize parameter
mu is too large? What happens when it is too small?
Exercise 13.15. Add (uncorrelated, normally distributed) noise into the simulation using the command r=filter(b,1,s)+sd*randn(size(s)). What is the
largest sd you can add, and still have no errors? Does the initial value for f
inuence this number? Try at least three initializations.
Exercise 13.16. Use DDequalizer.m to nd an equalizer that can open the eye
for the channel b=[1 1 -0.8 -0.3 1 1].
a. What equalizer length n is needed?
b. What initializations for f did you use?
c. How does the converged answer compare with the design in Exercises 13.3
and 13.9?
Exercise 13.17. Modify DDequalizer.m to generate a source sequence from the
alphabet 1, 3. For the default channel [0.5 1 -0.6], nd an equalizer that
opens the eye.
a. What equalizer length n is needed?
b. What initializations for f did you use?
c. Is this a fundamentally easier or more dicult task than when equalizing a
binary source?
d. How does the answer compare with the design in Exercises 13.4 and 13.10?
Section 13.6 provides the opportunity to view the simulated behavior of the
decision-directed equalizer, and to compare its performance with the other methods.

13.5

Dispersion-Minimizing Linear Equalization

This section considers an alternative performance function that leads to another
kind of blind equalizer. Observe that, for a binary 1 source, the square of the
source is known, even when the particular values of the source are not. Thus
s2 [k] = 1 for all k. This suggests creating a performance function that penalizes

291

Linear Equalization

Sampled
received
signal r[k]

y[k]
Equalizer

Adaptive
algorithm

X2

y2[k]

e[k]
Performance evaluation

Figure 13.8 A dispersion-minimizing adaptive linear equalizer for binary data uses the

dierence between the square of the received signal and unity to drive the adaptation
of the coecients of the equalizer.

the deviation from this known squared value = 1. In particular, consider

1
avg{( y 2 [k])2 },
4
which measures the dispersion of the equalizer output about its desired squared
value .
The associated adaptive element for updating the equalizer coecients is

JDM 
.
fi [k + 1] = fi [k]
fi fi =fi [k]
JDM =

Mimicking the derivation in (13.24) through (13.27) yields the dispersionminimizing algorithm5 (DMA) for blindly adapting the coecients of a linear
equalizer. The algorithm is
fi [k + 1] = fi [k] + avg{(1 y 2 [k])y[k]r[k i]}.
Suppressing the averaging operation, this becomes
fi [k + 1] = fi [k] + (1 y 2 [k])y[k]r[k i],

(13.29)

which is shown in the block diagram of Figure 13.8.

When the source alphabet is 1, then = 1. When the source is multilevel,
it is still useful to minimize the dispersion, but the constant should change to
= avg{s4 }/ avg{s2 }.
While DMA typically may converge to the desired answer from a worse initialization than decision-directed LMS, it is not as robust as trained LMS. For a
particular delay , the (average) squared recovery error surface being descended
(approximately) along the gradient by trained LMS is unimodal (i.e., it has
only one minimum). Therefore, no matter where the search is initialized, it nds
the desired sole minimum associated with the used in computing the source
recovery error. The dispersion performance function is multimodal with separate
minima corresponding to dierent achieved delays and polarities. To see this in
5

This is also known as the constant-modulus algorithm and as the Godard algorithm.

292

Chapter 13. Linear Equalization

the simplest case, observe that an answer in which all +1s are swapped with
all 1s has the same value at the optimal point. Thus, the convergent delay
and polarity achieved depend on the initialization used. A typical initialization
for DMA is a single nonzero spike located near the center of the equalizer. The
multimodal nature of DMA can be observed in the examples in the next section.
A simple Matlab program that implements the DMA algorithm is given in
DMAequalizer.m. The rst few lines dene the channel, create the binary source,
and pass the input through the channel. The last few lines implement the equalizer and calculate the error between the output of the equalizer and the source as
a way of measuring the performance of the equalizer. These parts of the code are
familiar from LSequalizer.m. The new part of the code is in the center, which
denes the length n of the equalizer, the stepsize mu of the algorithm, and the
initialization of the equalizer (which defaults to a center-spike initialization).
The coecients of the equalizer are updated as in (13.29). As with the other
equalizers, the code in EqualizerTest.m can be used to test the operation of
the converged equalizer.
Listing 13.5. DMAequalizer.m nd a DMA equalizer f for the channel b
b=[0.5 1 0.6];
m=1000; s=sign ( randn ( 1 ,m) ) ;
r=f i l t e r ( b , 1 , s ) ;
n=4; f =[0 1 0 0 ] ;
mu= . 0 1 ;
f o r i=n+1:m
r r=r ( i : 1 : i n + 1 ) ;
e =( f * r r ) * (1 ( f * r r ) 2 ) ;
f=f+mu* e * r r ;
end

%
%
%
%
%
%
%
%
%

def i ne channel
binary source of length m
output o f channel
c e n t e r s p i k e i n i t i a l i z a t i o n
algorithm s t e p s i z e
iterate
vector of received s i gnal
calculate error
update e q u a l i z e r c o e f f i c i e n t s

Running DMAequalizer.m results in an equalizer that is numerically similar

to the equalizers of the previous two sections. Initializing with the spike at different locations results in equalizers with dierent eective delays. The following
exercises are intended to encourage you to explore the DMA equalizer method.

Exercise 13.18. Try the initialization f=[0 0 0 0] in DMAequalizer.m. With

this initialization, can the algorithm open the eye? Try increasing m. Try changing
the stepsize mu. What other nonzero initializations will work?
Exercise 13.19. What happens in DMAequalizer.m when the stepsize parameter mu is too large? What happens when it is too small?
Exercise 13.20. Add (uncorrelated, normally distributed) noise into the simulation using the command r=filter(b,1,s)+sd*randn(size(s)). What is the
largest sd you can add and still have no errors? Does the initial value for f
inuence this number? Try at least three initializations.

293

Linear Equalization

Exercise 13.21. Use DMAequalizer.m to nd an equalizer that can open the

eye for the channel b=[1 1 -0.8 -0.3 1 1].
a. What equalizer length n is needed?
b. What initializations for f did you use?
c. How does the converged answer compare with the designs in Exercises 13.3,
13.9, and 13.16?
Exercise 13.22. Modify DMAequalizer.m to generate a source sequence from
the alphabet 1, 3. For the default channel [0.5 1 -0.6], nd an equalizer
that opens the eye.
a. What equalizer length n is needed?
b. What is an appropriate value of ?
c. What initializations for f did you use?
d. Is this a fundamentally easier or more dicult task than when equalizing a
binary source?
e. How does the answer compare with the designs in Exercises 13.4, 13.10,
and 13.17?
Exercise 13.23. Consider a DMA-like performance function J =
y 2 [k]|}. Show that the resulting gradient algorithm is

1
2

avg{|1

fi [k + 1] = fi [k] + avg{sign(1 y 2 [k])y[k]r[k i]}.

Hint: assume that the derivative of the absolute value is the sign function. Implement the algorithm and compare its performance with the DMA of (13.29) in
terms of
a. speed of convergence,
b. number of errors in a noisy environment (recall Exercise 13.20), and
c. ease of initialization.
Exercise 13.24. Consider a DMA-like performance function J = avg{|1
|y[k]||}. What is the resulting gradient algorithm? Implement your algorithm
and compare its performance with the DMA of (13.29) in terms of
a. speed of convergence of the equalizer coecients f,
b. number of errors in a noisy environment (recall Exercise 13.20), and
c. ease of initialization.

294

Chapter 13. Linear Equalization

13.6

Examples and Observations

This section uses the Matlab program dae.m which is available on the website.
The program demonstrates some of the properties of the least-squares solution to
the equalization problem and its adaptive cousins LMS, decision-directed LMS,
and DMA.6
The default settings in dae.m are used to implement the equalizer designs for
three channels. The source alphabet is a binary 1 signal. Each channel has an
FIR impulse response, and its output is summed with a sinusoidal interference
and some uniform white noise before reaching the receiver. The user is prompted
for the
1.
2.
3.
4.
5.
6.

choice of channels (0, 1, or 2),

maximum delay of the equalizer,
number of samples of training data,
gain of the sinusoidal interferer,
frequency of the sinusoidal interferer (in radians), and
magnitude of the white noise.

The program returns plots of the

1.
2.
3.
4.
5.
6.

received signal,
optimal equalizer output,
impulse response of the optimal equalizer and the channel,
recovery error at the output of the decision device,
zeros of the channel and the combined channelequalizer pair, and
magnitude and phase frequency responses of the channel, of the equalizer, and
of the combined channelequalizer pair.

For the default channels and values, these plots are shown in Figures 13.913.14.
the minimum average
T R,
The program also prints the condition number of R
squared recovery error (i.e., the minimum value achieved by the performance
function by the optimal equalizer for the optimum delay opt ), the optimal value
of the delay opt , and the percentage of decision device output errors in matching
the delayed source. These values were as follows:

r Channel 0

condition number: 130.2631

minimum value of performance function: 0.0534
optimum delay: 16
percentage of errors: 0

Throughout these simulations, other aspects of the system are assumed optimal; thus, the
downconversion is numerically perfect and the synchronization algorithms are assumed to
have attained their convergent values.

Linear Equalization

Received signal

295

Optimal equalizer output

2

1
1
0

1
2

1
0
1000 2000 3000 4000
Combined channel and optimal
equalizer impulse response

1000

2000

3000

4000

Decision device recovery error

1.2

0.8

0.5
0

0.4

0.5

0
0

10

20

30

40

1000

2000

3000

4000

Figure 13.9 Trained least-squares equalizer for channel 0: time responses. The received

signal is messy and cannot be used directly to recover the message. After passing
through the optimal equalizer, there is sucient separation to open the eye. The
bottom-left gure shows the impulse response of the channel convolved with the
impulse response of the optimal equalizer; it is close to an ideal response (which
would be unity at one delay and zero everywhere else). The bottom-right plot shows
that the message signal is recovered without error.

r Channel 1

condition number: 14.795

minimum value of performance function: 0.0307
optimum delay: 12
percentage of errors: 0
Channel 2
condition number: 164.1081
minimum value of performance function: 0.0300
optimum delay: 10
percentage of errors: 0

To see what these gures mean, consider the eight plots contained in Figures
13.9 and 13.10. The rst plot is the received signal, which contains the transmitted signal corrupted by the sinusoidal interferer and the white noise. After
the equalizer design, this received signal is passed through the equalizer, and the
output is shown in the plot entitled optimal equalizer output. The equalizer
transforms the data in the received signal into two horizontal stripes. Passing

Chapter 13. Linear Equalization

Freq Resp Magnitude

20

0.5

10
db

Imaginary part

FIR channel zeros

0
0.5
1

0
10

1 0.5 0 0.5
Real part

20

Zeros of channel/equalizer combination

1
0.5
0
-0.5
1
1

0
1
Real part

1
2
3
Normalized frequency

Freq Resp Phase

radians

Imaginary part

296

10
0
10
20
30
40
50
60

1
2
3
Normalized frequency

Figure 13.10 Trained least-squares equalizer for channel 0: singularities and frequency

responses. The large circles show the locations of the zeros of the channel in the
upper-left plot and the locations of the zeros of the combined channelequalizer pair
in the lower-left plot. The dotted line represents the frequency response of the
channel, the dashed line is the frequency response of the equalizer, and the solid line
is the frequency response of the combined channelequalizer pair.

Received signal
1

0.5

0.5

0
0.5

0.5
1

Optimal equalizer output

1.5

1
0

1000 2000 3000 4000

Combined channel and optimal

equalizer impulse response
1.2
1
0.8
0.6
0.4
0.2
0
0.2
0
10
20
30
40

1.5

1000 2000 3000 4000

Decision device recovery error

1
0.5
0
0.5
1

1000 2000 3000 4000

Figure 13.11 Trained least-squares equalizer for channel 1: time responses. As in

Figure 13.9, the equalizer is able to eectively undo the eects of the channel.

Linear Equalization

FIR channel zeros

297

Freq Resp Magnitude

1
0.5
0
0.5
1

10
db

Imaginary part

20

0
10

20

0
1
Real part

2
0
2
4
10 8 6 4 2 0
Real part

Freq Resp Phase

0

4
radians

Imaginary part

Zeros of channel/equalizer combination

1
2
3
4
Normalized frequency

10
20
30
40

1
2
3
4
Normalized frequency

Figure 13.12 Trained least-squares equalizer for channel 1: singularities and frequency

responses. The large circles show the locations of the zeros of the channel in the
upper-left plot and the locations of the zeros of the combined channelequalizer pair
in the lower-left plot. The dotted line represents the frequency response of the
channel, the dashed line is the frequency response of the equalizer, and the solid line
is the frequency response of the combined channelequalizer pair.

this through a simple sign device recovers the transmitted signal.7 The width of
these stripes is related to the cluster variance. The dierence between the sign of
the output of the equalizer and the transmitted data is shown in the plot labeled
decision device recovery error. This is zero, indicating that the equalizer has
done its job. The plot entitled combined channel and optimal equalizer impulse
response shows the convolution of the impulse response of the channel with the
impulse response of the equalizer. If the design was perfect and there was no
interference present, one tap of this combination would be unity and all the rest
would be zero. In this case, the actual design is close to this ideal.
The plots in Figure 13.10 show the same situation, but in the frequency
domain. The zeros of the channel are depicted in the upper-left plot. This constellation of zeros corresponds to the darkest of the frequency responses drawn
in the second plot. The primarily lowpass character of the channel can be intuited directly from the zero plot with the technique of Section F.2. The T -spaced
equalizer, accordingly, has a primarily highpass character, as can be seen from
the dashed frequency response in the upper-right plot of Figure 13.10. Combining these two gives the response in the middle. This middle response (plotted
with the solid line) is mostly at, except for a large dip at 1.4 radians. This
is exactly the frequency of the sinusoidal interferer, and this demonstrates the
7

Without the equalizer, the sign function would be applied directly to the received signal, and
the result would bear little relationship to the transmitted signal.

298

Chapter 13. Linear Equalization

Received signal

Optimal equalizer output

1.5
1
0.5
0

0.5
1
1.5

1000

2000

3000

4000

1.5
1
0.5
0
0.5
1
1.5
2

Combined channel and optimal

equalizer impulse response
1.2
1
0.8
0.6
0.4
0.2
0
0.2

1000

2000

3000

4000

Decision device recovery error

1
0.5
0

0.5
0

10

20

30

40

50

1000

2000

3000

4000

Figure 13.13 Trained least-squares equalizer for channel 2: time responses. Even for

this fairly severe channel, the equalizer is able to eectively undo the eects of the
channel as in Figures 13.9 and 13.11.

second major use of the equalizer; it is capable of removing uncorrelated interferences. Observe that the equalizer design is given no knowledge of the frequency
of the interference, nor even that any interference exists. Nonetheless, it automatically compensates for the narrowband interference by building a notch at
the oending frequency. The plot labeled Zeros of channel/equalizer combination shows the zeros of the convolution of the impulse response of the channel
and the impulse response of the optimal equalizer. Were the ring of zeros at
a uniform distance from the unit circle, then the magnitude of the frequency
response would be nearly at. But observe that one pair of zeros (at 1.4 radians) is considerably closer to the circle than all the others. Since the magnitude
of the frequency response is the product of the distances from the zeros to the
unit circle, this distance becomes small where the zero comes close. This causes
the notch.8
The eight plots for each of the other channels are displayed in similar fashion
in Figures 13.11 to 13.14.
Figures 13.1513.17 demonstrate equalizer design using the various recursive
methods of Sections 13.3 to 13.5 on the same problem. After running the leastsquares design in dae.m, the script asks whether you wish to simulate a recursive
solution. If yes, then you can choose

r which algorithm to run (trained LMS, decision-directed LMS, or blind DMA),

r the stepsize, and
8

If this kind of argument relating the zeros of the transfer function to the frequency response
of the system seems unfamiliar, see Appendix F.

Linear Equalization

30
20
10
0
10
20
30

0
1
Real part

radians

Zeros of channel/equalizer combination

1.5
1
0.5
0
0.5
1
1.5
0
1
2
1
Real part

Imaginary part

Freq Resp Magnitude

db

Imaginary part

FIR channel zeros

1.5
1
0.5
0
0.5
1
1.5

299

1
2
3
4
Normalized frequency
Freq Resp Phase

0
5
10
15
20
25
30
35

1
2
3
4
Normalized frequency

Figure 13.14 Trained least-squares equalizer for channel 2: singularities and frequency

responses. The large circles show the locations of the zeros of the channel in the
upper-left plot and the locations of the zeros of the combined channelequalizer pair
in the lower-left plot. The dotted line represents the frequency response of the
channel, the dashed line is the frequency response of the equalizer, and the solid line
is the frequency response of the combined channelequalizer pair.

r the initialization: a scale factor species the size of the ball about the optimum
equalizer within which the initial value for the equalizer is randomly chosen.
As is apparent from Figures 13.1513.17, all three adaptive schemes are successful with the recommended default values, which were used in equalizing
channel 0. All three exhibit, in the upper-left plots of Figures 13.1513.17, decaying averaged squared parameter error relative to their respective trained leastsquares equalizer for the data block. This means that all are converging to the
vicinity of the trained least-squares equalizer about which dae.m initializes the
algorithms. The collapse of the squared prediction error is apparent from the
upper-right plot in each of these gures. An initially closed eye appears for a
short while in each of the lower-left plots of equalizer output history in the gures. The match of the magnitudes of the frequency responses of the trained
(block) least-squares equalizer (plotted with the solid line) and the last adaptive
equalizer setting (plotted with asterisks) from the data-block stream is quite
striking in the lower-right plots in these gures.
As expected, we nd the following.

r With modest noise, as in the cases here outside the frequency band occupied
by the single narrowband interferer, the magnitude of the frequency response
of the trained least-squares solution exhibits peaks (valleys) where the channel
response has valleys (peaks) so that the combined response is nearly at. The
phase of the trained least-squares equalizer adds with the channel phase so

100

101

1000 2000 3000

Iterations

Squared prediction error

101

2
1.5
1
0.5
0

1000 2000 3000 4000

Iterations

0
1

10

2
3

3
2.5

Combined magnitude response

5

dB

Summed squared parameter error

Chapter 13. Linear Equalization

Adaptive equalizer output

300

1000 2000 3000 4000

Iterations

15

1
2
3
Normalized frequency

Figure 13.15 Trained LMS equalizer for channel 0. The dotted line represents the

achieved frequency response of the equalizer while the solid line represents the
frequency response of the desired (optimal) mean-square-error solution.

that their combination approximates a linear phase curve. Refer to plots in

the right columns of Figures 13.10, 13.12, and 13.14.
With modest channel noise and interferers, as the length of the equalizer
increases, the zeros of the combined channel and equalizer form rings. The
rings are denser the nearer the channel zeros are to the unit circle.

There are many ways in which the program dae.m can be used to investigate
and learn about equalization. Try to choose the various parameters to observe
the following
1. Increasing the power of the channel noise suppresses the frequency response of
the least-squares equalizer, with those frequency bands most suppressed being
those in which the channel has a null (and the equalizerwithout channel
noisewould have a peak).
2. Increasing the gain of a narrowband interferer results in a deepening of a
notch in the trained least-squares equalizer at the frequency of the interferer.
3. DMA is considered slower than trained LMS. Do you nd that DMA takes
longer to converge? Can you think of why it might be slower?
4. DMA typically accommodates larger initialization error than decision-directed
LMS. Can you nd cases where, with the same initialization, DMA converges
to an error-free solution but the decision-directed LMS does not? Do you think
there are cases in which the opposite holds?

100

101

1000 2000 3000

Iterations

Squared prediction error

101

2
1.5
1
0.5
0

1000 2000 3000 4000

Iterations

1
0
1

10

2
3

301

Combined magnitude response

5

dB

Adaptive equalizer output

Summed squared parameter error

Linear Equalization

1000 2000 3000 4000

Iterations

15

1
2
3
Normalized frequency

Figure 13.16 Decision-directed LMS equalizer for channel 0. The dotted line represents

101

100

1000 2000 3000

Iterations

4
3
2
1
0

1000 2000 3000 4000

Iterations

Combined magnitude response

5

2
1

0
5

dB

Adaptive equalizer output

101
0

Squared prediction error

Summed squared parameter error

the achieved frequency response of the equalizer while the solid line represents the
frequency response of the desired (optimal) mean-square-error solution.

0
1

10

2
3

1000 2000 3000 4000

Iterations

15

1
2
3
Normalized frequency

Figure 13.17 Blind DMA equalizer for channel 0. The dotted line represents the

achieved frequency response of the equalizer while the solid line represents the
frequency response of the desired (optimal) mean-square-error solution.

302

Chapter 13. Linear Equalization

5. It is necessary to specify the delay for the trained LMS, whereas the blind
methods do not require the parameter . Rather, the selection of an appropriate delay is implicit in the initialization of the equalizer coecients. Can
you nd a case in which, with the delay poorly specied, DMA outperforms
trained LMS from the same initialization?

For Further Reading

A comprehensive survey of trained adaptive equalization can be found in

r S. U. H. Qureshi, Adaptive Equalization, Proceedings of the IEEE, vol. 73,

no. 9, pp. 13491387, Sept. 1985.
An overview of the analytical tools that can be used to analyze LMS-style adaptive algorithms can be found in

r W. A. Sethares, The LMS Family, in Ecient System Identication and Signal Processing Algorithms, Eds. N. Kalouptsidis and S. Theodoridis, PrenticeHall, 1993.
A copy of this paper can also be found on the accompanying website.
One of our favorite discussions of adaptive methods is

r C. R. Johnson Jr., Lectures on Adaptive Parameter Estimation, Prentice-Hall,

1988.
This whole book can be found in .pdf form on the website.
An extensive discussion of equalization can also be found in Equalization on
the website.

14 Coding

The underlying purpose of any communication system is to transmit information.

But what exactly is information? How is it measured? Are there limits to the
amount of data that can be sent over a channel, even when all the parts of the
system are operating at their best? This chapter addresses these fundamental
questions using the ideas of Claude Shannon (19162001), who dened a measure
of information in terms of bits. The number of bits per second that can be
transmitted over the channel (taking into account its bandwidth, the power of
the signal, and the noise) is called the bit rate, and can be used to dene the
capacity of the channel.
Unfortunately, Shannons results do not give a recipe for how to construct
a system that achieves the optimal bit rate. Earlier chapters have highlighted
several problems that can arise in communication systems (including synchronization errors such as intersymbol interference). This chapter assumes that all
of these are perfectly mitigated. Thus, in Figure 14.1, the inner parts of the
communication system are assumed to be ideal, except for the presence of channel noise. Even so, most systems still fall far short of the optimal performance
promised by Shannon.
There are two problems. First, most messages that people want to send are
redundant, and the redundancy squanders the capacity of the channel. A solution is to preprocess the message so as to remove the redundancies. This is called
source coding, and is discussed in Section 14.5. For instance, as demonstrated in
Section 14.2, any natural language (such as English), whether spoken or written,
is repetitive. Information theory (as Shannons approach is called) quanties the
repetitiveness, and gives a way to judge the eciency of a source code by comparing the information content of the message with the number of bits required
by the code.
The second problem is that messages must be resistant to noise. If a message
arrives at the receiver in garbled form, then the system has failed. A solution is
to preprocess the message by adding extra bits, which can be used to determine
whether an error has occurred, and to correct errors when they do occur. For

304

Chapter 14. Coding

(Inherently redundant)
message to be
transmitted
Source
coding

Compressed
(redundancy
removed)

Structured
redundancy
added
Channel
coding

Transmitter

Structured
redundancy
removed

Noise
Channel

Receiver

Undo
channel
coding

Reconstructed
message
Undo
source
coding

Figure 14.1 Source and channel coding techniques manage redundancies in digital

communication systems: by rst removing inherent redundancies in the message, then

adding structured redundancies (which aid in combatting noise problems) and nally
restoring the original message.

example, one simple system would transmit each bit three times. Whenever a
single-bit error occurs in transmission, then the decoder at the receiver can gure
out by a simple voting rule that the error has occurred and what the bit should
have been. Schemes for nding and removing errors are called error-correcting
codes or channel codes, and are discussed in Section 14.6.
At rst glance, this appears paradoxical; source coding is used to remove
redundancy, while channel coding is used to add redundancy. But it is not really
self-defeating or contradictory because the redundancy that is removed by source
coding does not have a structure or pattern that a computer algorithm at the
receiver can exploit to detect or correct errors. The redundancy that is added in
channel coding is highly structured, and can be exploited by computer programs
implementing the appropriate decoding routines. Thus Figure 14.1 begins with
a message, and uses a source code to remove the redundancy. This is then coded
again by the channel encoder to add structured redundancy, and the resulting
signal provides the input to the transmitter of the previous chapters. One of the
triumphs of modern digital communications systems is that, by clever choice of
source and channel codes, it is possible to get close to the Shannon limits and
to utilize all the capacity of a channel.

14.1

What Is Information?
Like many common English words, information has many meanings. The American Heritage Dictionary catalogs six:
1.
2.
3.
4.

Knowledge derived from study, experience, or instruction.

Knowledge of a specic event or situation; intelligence.
A collection of facts or data.
The act of informing or the condition of being informed; communication of
knowledge.
5. Computer Science. A nonaccidental signal or character used as an input to a
computer or communication system.
6. A numerical measure of the uncertainty of an experimental outcome.

Coding

305

It would clearly be impossible to capture all of these senses in a technical denition that would be useful in transmission systems. The nal denition is closest
to our needs, though it does not specify exactly how the numerical measure
should be calculated. Shannon does. Shannons insight was that there is a simple relationship between the amount of information conveyed in a message and
the probability of the message being sent. This does not apply directly to messages such as sentences, images, or .wav les, but to the symbols of the alphabet
that are transmitted.
For instance, suppose that a fair coin has heads H on one side and tails T
on the other. The two outcomes are equally uncertain, and receiving either H
or T removes the same amount of uncertainty (conveys the same amount of
information). But suppose the coin is biased. The extreme case occurs when the
probability of H is 1. Then, when H is received, no information is conveyed,
because H is the only possible choice! Now suppose that the probability of sending H is 0.9 while the probability of sending T is 0.1. Then, if H is received, it
removes a little uncertainty, but not much. H is expected, since it usually occurs.
But if T is received, it is somewhat unusual, and hence conveys a lot of information. In general, events that occur with high probability give little information,
while events of low probability give considerable information.
To make this relationship between the probability of events and information
more plain, imagine a game in which you must guess a word chosen at random
from the dictionary. You are given the starting letter as a hint. If the hint is
that the rst letter is t, then this does not narrow down the possibilities very
much, since so many words start with t. But if the hint is that the rst letter
is x, then there are far fewer choices. The likely letter (the highly probable
t) conveys little information, while the unlikely letter (the improbable x)
conveys a lot more information by narrowing down the choices.
Heres another everyday example. Someone living in Ithaca (New York) would
be completely unsurprised that the weather forecast called for rain, and such
a prediction would convey little real information since it rains frequently. On
the other hand, to someone living in Reno (Nevada), a forecast of rain would
be very surprising, and would convey that very unusual meteorological events
were at hand. In short, it would convey considerable information. Again, the
amount of information conveyed is inversely proportional to the probabilities of
the events.
To transform this informal argument into a mathematical statement, consider
a set of N possible events xi , for i = 1, 2, . . . , N . Each event represents one possible outcome of an experiment, such as the ipping of a coin or the transmission
of a symbol across a communication channel. Let p(xi ) be the probability that
the ith event occurs, and suppose that some event must occur.1 This means that
1

When ipping the coin, it cannot roll into the corner and stand on its edge; each ip results
in either H or T .

306

Chapter 14. Coding

N

i=1 p(xi ) = 1. The goal is to nd a function I(xi ) that represents the amount
of information conveyed by each outcome.
The following three qualitative conditions all relate the probabilities of events
with the amount of information they convey:

(i) p(xi ) = p(xj ) I(xi ) = I(xj );

(ii) p(xi ) < p(xj ) I(xi ) > I(xj );
I(xi ) = 0.
(iii) p(xi ) = 1

(14.1)

Thus, receipt of the symbol xi should

1. give the same amount of information as receipt of xj if they are equally likely,
2. give more information if xi is less likely than xj , and
3. convey no information if it is known a priori that xi is the only alternative.
What kinds of functions I(xi ) fulll these requirements? There are many. For
instance, I(xi ) = 1/p(xi ) 1 and I(xi ) = (1 p2 (xi ))/p(xi ) both fulll (i)(iii).
To narrow down the possibilities, consider what happens when a series of
experiments is conducted, or equivalently, when a series of symbols is transmitted. Intuitively, it seems reasonable that, if xi occurs at one trial and xj occurs
at the next, the total information in the two trials should be the sum of the
information conveyed by receipt of xi and the information conveyed by receipt
of xj ; that is, I(xi ) + I(xj ). This assumes that the two trials are independent
of each other, namely, that the second trial is not inuenced by the outcome of
rst (and vice versa).
Formally, two events are dened to be independent if the probability that both
occur is equal to the product of the individual probabilitiesthat is, if
p(xi and xj ) = p(xi )p(xj ),

(14.2)

where p(xi and xj ) means that xi occurred in the rst trial and xj occurred
in the second. This additivity requirement for the amount of information conveyed by the occurrence of independent events is formally stated in terms of the
information function as
(iv) I(xi and xj ) = I(xi ) + I(xj )
when the events xi and xj are independent.
On combining the additivity in (iv) with the three conditions (i)(iii), there is
one (and only one) possibility for I(xi ):


1
= log(p(xi )).
I(xi ) = log
(14.3)
p(xi )

Coding

307

It is easy to see that (i)(iii) are fullled, and (iv) follows from the properties of
the log (recall that log(ab) = log(a) + log(b)). Therefore,




1
1
I(xi and xj ) = log
= log
p(xi and xj )
p(xi )p(xj )




1
1
+ log
= log
p(xi )
p(xj )
= I(xi ) + I(xj ).
The base of the logarithm can be any (positive) number. The most common
choice is base 2, in which case the measurement of information is called bits.
Unless otherwise stated explicitly, all logs in this chapter are assumed to be
base 2.

Example 14.1. Suppose there are N = 3 symbols in the alphabet, which are
transmitted with probabilities p(x1 ) = 1/2, p(x2 ) = 1/4, and p(x3 ) = 1/4. Then
the information conveyed by receiving x1 is 1 bit, since


1
= log(2) = 1.
I(x1 ) = log
p(x1 )
Similarly, the information conveyed by receiving either x2 or x3 is I(x2 ) =
I(x3 ) = log(4) = 2 bits.

Example 14.2. Suppose that a length m binary sequence is transmitted, with

all symbols equally probable. Thus N = 2m , xi is the binary representation of
the ith symbol for i = 1, 2, . . . , N , and p(xi ) = 2m . The information contained
in the receipt of any given symbol is


1
I(xi ) = log
= log(2m ) = m bits.
p(xi )

Exercise 14.1. Consider a standard six-sided die. Identify N , xi , and p(xi ).

How many bits of information are conveyed if a 3 is rolled? Now roll two dice,
and suppose the total is 12. How many bits of information does this represent?
Exercise 14.2. Consider transmitting a signal with values chosen from the sixlevel alphabet 1, 3, 5.
a. Suppose that all six symbols are equally likely. Identify N , xi , and p(xi ), and
calculate the information I(xi ) associated with each i.
b. Suppose instead that the symbols 1 occur with probability 1/4 each, 3
occur with probability 1/8 each, and 5 occurs with probability 1/4. What

308

Chapter 14. Coding

percentage of the time is 5 transmitted? What is the information conveyed

by each of the symbols?
Exercise 14.3. The 8-bit binary ASCII representation of any letter (or
any character of the keyboard) can be found using the Matlab command
dec2bin(text), where text is any string. Using ASCII, how much information
is contained in the letter a, assuming that all the letters are equally probable?
Exercise 14.4. Consider a decimal representation of = 3.1415926 . . . Calculate the information (number of bits) required to transmit successive digits of ,
assuming that the digits are independent. Identify N , xi , and p(xi ). How much
information is contained in the rst million digits of ?
There is an alternative denition of information (in common usage in the
mathematical-logic and computer-science communities) in which information is
dened in terms of the complexity of representation, rather than in terms of
the reduction in uncertainty. Informally speaking, this alternative denes the
complexity (or information content) of a message by the length of the shortest
computer program that can replicate the message. For many kinds of data, such
as a sequence of random numbers, the two measures agree because the shortest
program that can represent the sequence is just a listing of the sequence. But in
other cases, they can dier dramatically. Consider transmitting the rst million
digits of the number . Shannons denition gives a large information content (as
in Exercise 14.4), while the complete sequence can, in principle, be transmitted
with a very short computer program.

14.2

Redundancy
All the examples in the previous section presume that there is no relationship
between successive symbols. (This was the independence assumption in (14.2).)
This section shows by example that real messages often have signicant correlation between symbols, which is a kind of redundancy. Consider the following
sentence from Shannons paper A Mathematical Theory of Communication:
It is clear, however, that by sending the
information in a redundant form the
probability of errors can be reduced.
This sentence contains 20 words and 115 characters, including the commas,
period, and spaces. It can be coded into the 8-bit binary ASCII character
set recognized by computers as the text format, which translates the character string (that is readable by humans) into a binary string containing 920
(= 8 115) bits.

Coding

309

Suppose that Shannons sentence is transmitted, but that errors occur so that
1% of the bits are ipped from one to zero (or from zero to one). Then about
3.5% of the letters have errors:
It is clea2, however, that by sendhng the
information in a redundaNt form the
probabilipy of errors can be reduced.
The message is comprehensible, although it appears to have been typed poorly.
With 2% bit error, about 7% of the letters have errors:
It is clear, howaver, thad by sending the
information in a redundan4 form phe
prkbability of errors cAf be reduced.
Still the underlying meaning is decipherable. A dedicated reader can often decipher text with up to about 3% bit error (10% symbol error). Thus, the message
has been conveyed, despite the presence of the errors. The reader, with an extensive familiarity with English words, sentences, and syntax, is able to recognize
the presence of the errors and to correct them.
As the bit error rate grows to 10%, about one third of the letters have errors,
and many words have become incomprehensible. Because space is represented
as an ASCII character just like all the other symbols, errors can transform spaces
into letters or letters into spaces, thus blurring the true boundaries between the
words.
Wt is ahear, h/wav3p, dhat by sending phc
)hformatIon if a rEdundaft fnre thd
prkba@)hity ob erropc can be reduaed.
With 20% bit error, about half of the letters have errors and the message is
completely illegible:
I4 "s Cdrq h+Ae&d"( (At by sjdafd th$
hfFoPmati/. )f a p(d5jdan fLbe thd
rab!DITy o& drkp1 aa& bE rd@u!ed.
These examples were all generated using the following Matlab program
redundant.m which takes the text textm, translates it into a binary string, and
then causes per percent of the bits to be ipped. The program then gathers
statistics on the resulting numbers of bit errors and symbol errors (how many
letters were changed).
Listing 14.1. redundant.m redundancy of written english in bits and letters
textm= It is clear , however , that by sending the ...

310

Chapter 14. Coding

informati o n in a redundant form the ...

probabili t y of errors can be reduced .
% 8 b i t a s c i i ( b i n a r y ) e q u i v a l e n t o f t e x t
ascm=d e c 2 b i n ( d o u b l e ( textm ) , 8 ) ;
% tu r n i n t o one l o n g b i n a r y s t r i n g
binm=reshape ( ascm , 1 , 8 * length ( textm ) ) ;
per =.01;
% pr obabi l i ty of bit error
f o r i =1:8 * length ( textm )
r=rand ;
% swap 0 and 1 with p r o b a b i l i t y p er
i f ( r>1p e r ) & binm ( i )==0 , binm ( i )=1 ; end
i f ( r>1p e r ) & binm ( i )==1 , binm ( i )=0 ; end
end
a s c r=reshape ( binm , 8 , length ( textm ) ) ;
% to a s c i i b i n a r y
t e x t r=s e t s t r ( b i n 2 d e c ( a s c r ) )
% to t e x t
b i t e r r o r=sum(sum( abs ( a s c r ascm ) ) )
% # of bit er r or s
s y m e r r r o r=sum( sign ( abs ( textmt e x t r ) ) )
% # o f symbol e r r o r s
numwords=sum( sign ( find ( textm ==32)))+1
% # o f words i n textm
l e t t e r r o r=s y m e r r r o r / length ( textm )
% # of l e t t e r er r or s

Exercise 14.5. Read in a large text le using the following Matlab code. (Use
one of your own or use one of the included text les.)2 Make a plot of the symbol
error rate as a function of the bit error rate by running redundant.m for a variety
of values of per. Examine the resulting text. At what value of per does the text
become unreadable? What is the corresponding symbol error rate?
Listing 14.2. readtext.m read in a text document and translate to character string
[ f i d , m e s s a g e i ] = fopen ( OZ . txt , r ) ;
f d a t a=fread ( f i d ) ;
text=c h a r ( f d a t a ) ;

% f i l e must be t e x t
% r ea d t e x t a s a v e c t o r
% to c h a r a c t e r s t r i n g

Thus, for English text encoded as ASCII characters, a signicant number of

errors can occur (about 10% of the letters can be arbitrarily changed) without
altering the meaning of the sentence. While these kinds of errors can be corrected
by a human reader, the redundancy is not in a form that is easily exploited by a
computer. Even imagining that the computer could look up words in a dictionary,
the person knows from context that It is clear is a more likely phrase than It
is clean when correcting Shannons sentence with 1% errors. The person can
gure out from context that cAf (from the phrase with 2% bit errors) must
have had two errors by using the long-term correlation of the sentence (i.e., its
meaning). Computers do not deal readily with meaning.3
In the previous section, the information contained in a message was dened to
depend on two factors: the number of symbols and their probability of occurrence.
But this assumes that the symbols do not interactthat the letters are indepen2
3

Through the Looking Glass by Lewis Carroll (carroll.txt) and The Wonderful Wizard of Oz
by Frank Baum (OZ.txt) are available on the website.
A more optimistic rendering of this sentence: Computers do not yet deal readily with meaning.

Coding

311

dent. How good an assumption is this for English text? It is a poor assumption.
As the preceding examples suggest, normal English is highly correlated.
It is easy to catalog the frequency of occurrence of the letters. The letter e
is the most common. In Frank Baums Wizard of Oz, for instance, e appears
20,345 times and t appears 14,811 times, but the letters q and x appear
only 131 and 139 times, respectively (z might be a bit more common in Baums
book than normal because of the title). The percentage occurrence for each letter
in the Wizard of Oz is
a
b
c
d
e
f
g

6.47
1.09
1.77
4.19
10.29
1.61
1.60

h
i
j
k
l
m
n

5.75
4.63
0.08
0.90
3.42
1.78
4.90

o
p
q
r
s
t
u

6.49
1.01
0.07
4.71
4.51
7.49
2.05

v
w
x
y
z

0.59
2.49
0.07
1.91
0.13

(14.4)

Space is the most frequent character, occurring 20% of the time. It was easier
to use the following Matlab code, in conjunction with readtext.m, than to
count the letters by hand.
Listing 14.3. freqtext.m frequency of occurrence of letters in text
l i t t l e =length ( find ( t e x t s t r i n g==t ) ) ;
b i g=length ( find ( t e x t s t r i n g==T ) ) ;
f r e q =( l i t t l e +b i g )/ length ( t e x t s t r i n g )

% how many t i m e s t o c c u r s
% how many t i m e s T o c c u r s
% percentage

If English letters were truly independent, then it should be possible to generate

English-like text using this table of probabilities. Here is a sample:
Od m shous t ad schthewe be amalllingod
ongoutorend youne he Any bupecape tsooa w
beves p le t ke teml ley une weg rloknd
which does not look anything like English. How can the nonindependence of the
text be modeled? One way is to consider the probabilities of successive pairs of
letters instead of the probabilities of individual letters. For instance, the pair
th is quite frequent, occurring 11,014 times in the Wizard of Oz, while sh
occurs 861 times. Unlikely pairs such as wd occur in only ve places4 and pk
not at all. For example, suppose that He was chosen rst. The next pair would
be e followed by something, with the probability of the something dictated by
the entries in the table. Following this procedure results in output like this:
Her gethe womfor if you the to had the sed
th and the wention At th youg the yout by
and a pow eve cank i as saing paill
4

In the words crowd and sawdust.

312

Chapter 14. Coding

Observe that most of the two-letter combinations are actual words, as well as
many three-letter words. Longer sets of symbols tend to wander improbably.
While, in principle, it would be possible to continue gathering probabilities of all
three-letter combinations, then four, etc., the table begins to get rather large (a
matrix with 26n elements would be needed to store all the n-letter probabilities).
Shannon5 suggests another way:
. . . one opens a book at random and selects a letter on the page. This letter is recorded.
The book is then opened to another page, and one reads until this letter is encountered.
The succeeding letter is then recorded. Turning to another page, this second letter is
searched for, and the succeeding letter recorded, etc.

Of course, Shannon did not have access to Matlab when he was writing in
1948. If he had, he might have written a program like textsim.m, which allows
specication of any text (with the default being The Wizard of Oz) and any number of terms for the probabilities. For instance, with m=1, the letters are chosen
completely independently; with m=2, the letters are chosen from successive pairs;
and with m=3, they are chosen from successive triplets. Thus, the probabilities
of clusters of letters are dened implicitly by the choice of the source text.
Listing 14.4. textsim.m use (large) text to simulate transition probabilities
m=1;
l i n e l e n g t h =60;
load OZ . mat
n=text ( 1 :m) ; n l i n e=n ; n l e t=x ;
f o r i =1:100
j =1;
while j <l i n e l e n g t h | n l e t =
k=f i n d s t r ( text , n ) ;
i n d=round ( ( length ( k ) 1) * rand )+1;
n l e t=text ( k ( i n d )+m) ;
i f abs ( n l e t )==13
n l e t= ;
end
n l i n e =[ n l i n e , n l e t ] ;
n=[n ( 2 :m) , n l e t ] ;
j=j +1;
end
disp ( [ n l i n e s e t s t r ( 1 3 ) ] )
n l i n e= ;
end

%
%
%
%
%

# terms f o r t r a n s i t i o n
# l e t t e r s i n each l i n e
f i l e f o r input
i n i t i a l i z e variables
# l i n e s i n o u tp u t

% s c a n th r o u g h f i l e
% a l l occurrences of seed
% p i c k one
% treat carriage returns
% as s p aces
% add n e x t l e t t e r
% new s e e d

% f o r m a t o u t p u t / add CRs
% i n i t i a l i z e next l i n e

Typical output of textsim.m depends heavily on the number of terms m used for
the transition probabilities. With m=1 or m=2, the results appear much as above.
When m=3,
Be end claime armed yes a bigged wenty for me
fearabbag girl Humagine ther mightmarkling the
5

C. E. Shannon, A Mathematical Theory of Communication, The Bell System Technical

Journal, vol. 27, nos. 7 and 10, pp. 379423 and 623656, July and October, 1948.

Coding

313

many the scarecrow pass and I havely and lovery

wine end at then only we pure never
many words appear, and many combinations of letters that might be words,
but arent quite. Humagine is suggestive, though it is not clear exactly what
mightmarkling might mean. When m=4,
Water of everythinkies friends of the scarecrow
no head we She time toto be well as some
although to they would her been Them became the
small directions and have a thing woodman
the vast majority of words are actual English, though the occasional conjunction
of words (such as everythinkies) is not uncommon. The output also begins to
strongly reect the text used to derive the probabilities. Since many four-letter
combinations occur only once, there is no choice for the method to continue
spelling a longer word; this is why the scarecrow and the woodman gure
prominently. For m=5 and above, the random output is recognizably English,
and strongly dependent on the text used:
Four trouble and to taken until the bread
hastened from its Back to you over the emerald
city and her in toward the will Trodden and
being she could soon and talk to travely lady

Exercise 14.6. Run the program textsim.m using the input le carroll.mat,
which contains the text to Lewis Carrolls Through the Looking Glass, with m =
1, 2, . . ., 8. At what point does the output repeat large phrases from the input
text?
Exercise 14.7. Run the program textsim.m using the input le foreign.mat,
which contains a book that is not in English. Looking at the output for various
m, can you tell what language the input is? What is the smallest m (if any) at
which it becomes obvious?
The following two problems might not appeal to everyone.
Exercise 14.8. The program textsim.m operates at the level of letters and the
probabilities of transition between successive sets of m-length letter sequences.
Write an analogous program that operates at the level of words and the probabilities of transition between successive sets of m-length word sequences. Does
your program generate plausible-sounding phrases or sentences?

314

Chapter 14. Coding

Exercise 14.9. There is nothing about the technique of textsim.m that is

inherently limited to dealing with text sequences. Consider a piece of (notated)
music as a sequence of symbols, labeled so that each C note is 1, each C  note
is 2, each D note is 3, etc. Create a table of transition probabilities from a piece
of music, and then generate new melodies in the same way that textsim.m
generates new sentences. (Observe that this procedure can be automated using
standard MIDI les as input.)
Because this method derives the multiletter probabilities directly from a text,
there is no need to compile transition probabilities for other languages. Using
Vergils Aeneid (with m=3) gives
Aenere omnibus praeviscrimus habes ergemio nam
inquae enies Media tibi troius antis igna volae
subilius ipsis dardatuli Cae sanguina fugis
ampora auso magnum patrix quis ait longuin
which is not real Latin. Similarly,
Que todose remosotro enga tendo en guinada y
ase aunque lo Se dicielos escubra la no fuerta
pare la paragales posa derse Y quija con figual
se don que espedios tras tu pales del
is not Spanish (the input le was Cervantes Don Quijote, also with m=3), and
Seule sontagne trait homarcher de la t au onze
le quance matices Maississait passepart penaient
la ples les au cherche de je Chamain peut accide
bien avaien rie se vent puis il nez pande
is not French (the source was Le Tour du Monde en Quatre Vingts Jours, a
translation of Jules Vernes Around the World in Eighty Days.)
The input le to the program textsim.m is a Matlab .mat le that has been
preprocessed to remove excessive line breaks, spaces, and capitalization using
textman.m, which is why there is no punctuation in these examples. A large
assortment of text les is available for downloading at the website of Project
Gutenberg (at http://www.gutenberg.org/).
Text, in a variety of languages, retains some of the character of its language
with correlations of 3 to 5 letters (2135 bits, when coded in ASCII). Thus, messages written in those languages are not independent, except possibly at lengths
greater than this. A result from probability theory suggests that, if the letters are
clustered into blocks that are longer than the correlation, then the blocks may
be (nearly) independent. This is one strategy to pursue when designing codes

Coding

315

that seek to optimize performance. Section 14.5 will explore some practical ways
to attack this problem, but the next two sections establish a measure of performance such that it is possible to know how close to the optimal any given code
lies.

14.3

Entropy
This section extends the concept of information from a single symbol to a
sequence of symbols. As dened by Shannon,6 the information in a symbol is
inversely proportional to its probability of occurring. Since messages are composed of sequences of symbols, it is important to be able to talk concretely about
the average ow of information. This is called the entropy and is formally dened
as
H(x) =

N


p(xi )I(xi )

i=1

N



p(xi ) log

i=1

N


1
p(xi )

p(xi ) log(p(xi )),

(14.5)

i=1

where the symbols are drawn from an alphabet xi , each with probability p(xi ).
H(x) sums the information in each symbol, weighted by the probability of that
symbol. Those familiar with probability and random variables will recognize
this as an expectation. Entropy7 is measured in bits per symbol, and so gives
a measure of the average amount of information transmitted by the symbols of
the source. Sources with dierent symbol sets and dierent probabilities have
dierent entropies. When the probabilities are known, the denition is easy to
apply.

Example 14.3. Consider the N = 3 symbol set dened in Example 14.1. The
entropy is
H(x) =

6
7

1
1
1
1 + 2 + 2 = 1.5 bits/symbol.
2
4
4

Actually, Hartley was the rst to use this as a measure of information in his 1928 paper in
the Bell System Technical Journal called Transmission of Information.
Warning: though the word is the same, this is not the same as the notion of entropy that is
familiar from physics since the units here are in bits per symbol while the units in physics
are energy per degree Kelvin.

316

Chapter 14. Coding

Exercise 14.10. Reconsider the fair die of Exercise 14.1. What is its entropy?

Example 14.4. Suppose that the message {x1 , x3 , x2 , x1 } is received from a

source characterized as follows.
1. N = 4, p(x1 ) = 0.5, p(x2 ) = 0.25, p(x3 ) = p(x4 ) = 0.125. The total information is


4
4


1
= 1 + 2 + 3 + 3 = 9 bits.
I(xi ) =
log
p(xi )
i=1
i=1
The entropy of the source is
1
1
1
1
1 + 2 + 3 + 3 = 1.75 bits/symbol.
2
4
8
8
2. N = 4, p(xi ) = 0.25 for all i. The total information is I(xi ) = 2 + 2 + 2 + 2 =
8. The entropy of the source is
H(x) =

1
1
1
1
2 + 2 + 2 + 2 = 2 bits/symbol.
4
4
4
4
In Example 14.4, messages of the same length from the rst source give less
information than those from the second source. Hence, sources with the same
number of symbols but dierent probabilities can have dierent entropies. The
key is to design a system to maximize entropy since this will have the largest
throughput, or largest average ow of information. But how can this be achieved?
First, consider the simple case in which there are two symbols in the alphabet,
x1 with probability p, and x2 with probability 1 p. (Think of a coin that is
weighted so as to give heads with higher probability than tails.) Applying the
denition (14.5) shows that the entropy is
H(x) =

H(p) = p log(p) (1 p) log(1 p).

This is plotted as a function of p in Figure 14.2. For all allowable values of
p, H(p) is positive. As p approaches either zero or one, H(p) approaches zero,
which represent the symmetric cases in which either x1 occurs all the time or x2
occurs all the time, and no information is conveyed. H(p) reaches its maximum
in the middle, at p = 0.5. For this example, entropy is maximized when the two
symbols are equally likely.
Exercise 14.11. Show that H(p) is maximized at p = 0.5 by taking the derivative and setting it equal to zero.
The next result shows that an N -symbol source cannot have entropy larger
than log(N ), and that this bound is achieved when all the symbols are equally
likely. Mathematically, H(x) log(N ), which is demonstrated by showing that

Coding

317

H(p)

0.8
0.6

Figure 14.2 Entropy of a binary

signal with probabilities p and
1 p.

0.4
0.2
0

0.1

0.2

0.3

0.4 0.5 0.6

Probability p

0.7

0.8

0.9

H(x) log(N ) 0. First,

H(x) log(N ) =

N



p(xi ) log

i=1

=
since

N
i=1

N



p(xi ) log

i=1

1
p(xi )
1
p(xi )


log(N )


N


p(xi ) log(N ),

i=1

p(xi ) = 1. Gathering terms, this can be rewritten

 


N

1
p(xi ) log
H(x) log(N ) =
log(N )
p(xi )
i=1


N

1
,
=
p(xi ) log
N p(xi )
i=1

and changing the base of the logarithm (using log(z) log2 (z) = log2 (e) ln(z),
where ln(z) loge (z)), gives


N

1
p(xi ) ln
H(x) log(N ) = log(e)
.
N p(xi )
i=1
If all symbols
are equally likely, p(xi ) = 1/N , then 1/(N p(xi )) = 1 and

1
=
ln(1)
= 0. Hence H(x) = log(N ). On the other hand, if the symln N p(x
i)
bols are not equally likely, then the inequality ln(z) z 1 (which holds for
z 0) implies that
N


1
1
N p(xi )
i=1
$N
%
N
 1


= log(e)
p(xi )
N
i=1
i=1

H(x) log(N ) log(e)

p(xi )

= log(e) [1 1] = 0.

(14.6)

Rearranging (14.6) gives the desired bound on the entropy, namely that H(x)
log(N ). This says that, all else being equal, it is preferable to choose a code

318

Chapter 14. Coding

in which each symbol occurs with the same probability. Indeed, Example 14.5
provides a concrete source for which the equal-probability case has higher entropy
than the unequal-probability case.
Section 14.2 showed how letters in the text of natural languages do not occur
with equal probability. Thus, naively using the letters will not lead to an ecient transmission. Rather, the letters must be carefully translated into equally
probable symbols in order to increase the entropy. A method for accomplishing
this translation is given in Section 14.5, but Section 14.4 examines the limits of
attainable performance when transmitting symbols across a noisy (but otherwise
perfect) channel.

14.4

Channel Capacity
Section 14.1 showed how much information (measured in bits) is contained in
a given symbol, and Section 14.3 generalized this to the average amount of
information contained in a sequence or set of symbols (measured in bits per
symbol). In order to be useful in a communication system, however, the data must
move from one place to another. What is the maximum amount of information
that can pass through a channel in a given amount of time? The main result of
this section is that the capacity of the channel denes the maximum possible ow
of information through the channel. The capacity is a function of the bandwidth
of the channel and of the amount of noise in the system, and it is measured in
bits per second.
If the data are encoded using N = 2 equally probable bits per symbol, and if
the symbols are independent, the entropy is H2 = 0.5 log(2) + 0.5 log(2) = 1 bit
per symbol. Why not increase the number of bits per symbol? This would allow
representing more information. On doubling to N = 4, the entropy increases to
H4 = 2. In general, when using N bits, the entropy is HN = log(N ). By increasing N without bound, the entropy can be increased without bound! But is it
really possible to send an innite amount of information?
When doubling the size of N , one of two things must happen. Either the distance between the levels must decrease, or the power must increase. For instance,
it is common to represent the binary signal as 1 and the four-level signal as
1, 3. In this representation, the distance between neighboring values is constant, but the power in the signal has increased. Recall that the power in a
discrete signal x[k] is
T
1  2
lim
x [k].
T T
k=1

For a binary signal with equal probabilities, this is P2 = 12 (12 + (1)2 ) = 1. The
four-level signal has power P4 = 14 (12 + (1)2 + 32 + (3)2 ) = 5. To normalize
the power to unity for the four-level signal, calculate the value x such that

Coding

319

N-level signals with equal power

2
1.5

Amplitude

1
0.5

Figure 14.3 When the power is

equal, the values of the
N -level signal grow closer as
N increases.

0
0.5
1
1.5
2
2-level signal 4-level signal 8-level signal 16-level signal

*
+ (x)2 + (3x)2 + (3x)2 ) = 1, which is x = 1/5. Figure 14.3 shows how
the values of the N -level signal become closer together as N increases, when the
power is held constant.
Now it will be clearer why it is not really possible to send an innite amount
of information in a single symbol. For a given transmitter power, the amplitudes
become closer together for large N , and the sensitivity to noise increases. Thus,
when there is noise (and some is inevitable), the four-level signal is more prone to
errors than the two-level signal. Said another way, a higher signal-to-noise ratio8
(SNR) is needed to maintain the same probability of error in the four-level signal
as in the two-level signal.
Consider the situation in terms of the bandwidth required to transmit a given
set of data containing M bits of information. From the Nyquist sampling theorem of Section 6.1, data can be sent through a channel of bandwidth B at a
maximum rate of 2B symbols per second. If these symbols are coded into two
levels, then M symbols must be sent. If the data are transmitted with four levels
(by assigning pairs of binary digits to each four-level symbol), then only M/2
symbols are required. Thus the multilevel signal can operate at half the symbol
rate of the binary signal. Said another way, the four-level signal requires only
half the bandwidth of the two-level signal.
The previous two paragraphs show the trade-o between SNR and bandwidth.
To maintain the same probability of error, larger bandwidth allows smaller SNR;
larger SNR allows the use of a narrower frequency band. Quantifying this tradeo was one of Shannons greatest contributions.
While the details of a formal proof of the channel capacity are complex, the
result is believable when thought of in terms of the relationship between the
1
2
4 (x

As the term suggests, SNR is the ratio of the energy (or power) in the signal to the energy
(or power) in the noise.

320

Chapter 14. Coding

distance between the levels in a source alphabet and the average amount of
noise that the system can tolerate. A digital signal with N levels has a maximum
information rate C = log(N )/T , where T is the time interval between transmitted
symbols. C is the capacity of the channel, and has units of bits per second.
This can be expressed in terms of the bandwidth B of the channel by recalling
Nyquists sampling theorem, which says that a maximum of 2B pulses per second
can pass through the channel. Thus the capacity can be rewritten
C = 2B log(N ) bits per second.
To include the eect of noise, observe that the power of the received signal is
S + P (where S is the power of the signal and P is the power
of the noise).
Accordingly, the average amplitude of the received signal is S + P and the
average amplitude of the noise is P. The average distance d between levels
is twice the
average amplitude divided by the number of levels (minus one),
and so d = 2 S + P/(N 1). Many errors will occur in the transmission unless
the distance between the signal levels is separated by at least twice the average
amplitude of the noise, that is, unless

2 S +P
> 2 P.
d=
N 1

Rearranging this implies that N 1 must be

no largerthan S + P/ P. The
actual bound (as Shannon shows) is that N S + P/ P, and using this value
gives




S
S +P

= B log 1 +
(14.7)
C = 2B log
P
P
bits per second.
Observe that, if either the bandwidth or the SNR is increased, so is the channel capacity. For white noise, as the bandwidth increases, the power in the noise
increases, the SNR decreases, and so the channel capacity does not become innite. For a xed channel capacity, it is easy to trade o bandwidth against SNR.
For example, suppose a capacity of 1000 bits per second is required. Using a
bandwidth of 1 kHz, we nd that the signal and the noise can be of equal power.
As the allowed bandwidth is decreased, the ratio S/P increases rapidly:
Bandwidth
1000
500
250
125
100

Hz
Hz
Hz
Hz
Hz

S/P
1
3
15
255
1023

Shannons result can now be stated succinctly. Suppose that there is a source
producing information at a rate of R bits per second and a channel of capacity
C. If R < C (where C is dened as in (14.7)) then there exists a way to represent

Coding

321

(or code) the data so that it can be transmitted with arbitrarily small error.
Otherwise, the probability of error is strictly positive.
This is tantalizing and frustrating at the same time. The channel capacity
denes the ultimate goal beyond which transmission systems cannot go, yet it
provides no recipe for how to achieve the goal. The next sections describe various
methods of representing or coding the data that assist in approaching this limit
in practice.
The following Matlab program explores a noisy system. A sequence of fourlevel data is generated by calling the pam.m routine. Noise is then added with
power specied by p, and the number of errors caused by this amount of noise
is calculated in err.
Listing 14.5. noisychan.m generate 4-level data and add noise
m=1000;
p =1/15; s = 1 . 0 ;
x=pam (m, 4 , s ) ;
L=sqrt ( 1 / 5 ) ;
n=sqrt ( p ) * randn ( 1 ,m) ;
y=x+n ;
qy=q u a n t a l p h ( y , [ 3 * L, L , L , 3 * L ] ) ;
e r r=sum( abs ( sign ( qy x ) ) ) /m;

%
%
%
%
%
%
%
%

l e n g t h o f data s e q u e n c e
power o f n o i s e and s i g n a l
4PAM i n p u t with power 1 . . .
. . . with amp l e v e l s L
n o i s e with power p
o u t p u t adds n o i s e to data
q u a n t i z e to [ 3 * L, L , L , 3 * L ]
percent transmission er r or s

Typical outputs of noisychan.m are shown in Figure 14.4. Each plot shows the
input sequence (the four solid horizontal lines), the input plus the noise (the cloud
of small dots), and the error between the input and quantized output (the dark
stars). Thus the dark stars that are not at zero represent errors in transmission.
The noise P in the right-hand case is the maximum noise allowable in the
plausibility argument used to derive (14.7), which relates the average amplitudes
of the signal plus the noise to the number of levels in the signal. For S = 1 (the
same conditions as in Exercise 14.12(a)), the noise was chosento be independent

and normally distributed with power P to ensure that 4 = 1 + P / P . The

middle plot used a noise with power P /3 and the left-hand plot had noise
power P /6. As can be seen from the plots, there were essentially no errors when
using the smallest noise, a handful of errors in the middle, and about 6% errors
when the power of the noise matched the Shannon capacity. Thus, this naive
transmission of four-level data (i.e., with no coding) has many more errors than
the Shannon limit suggests.
Exercise 14.12. Find the amplitudes of the N -level (equally spaced) signal with
unity power when
a. N = 4,
b. N = 6,
c. N = 8.
Exercise 14.13. Use noisychan.m to compare the noise performance of twolevel, four-level, and six-level transmissions.

Chapter 14. Coding

1
Amplitude

322

2000 4000
1/6 max noise

2000 4000
1/3 max noise

2000 4000
max noise

Figure 14.4 Each plot shows a four-level PAM signal (the four solid lines), the signal

plus noise (the scattered dots), and the error between the data and the quantized
output
stars). The noise in the right-hand plot was at the Shannon limit
(the dark
N S + P / P , the noise in the middle plot was at one-third the power, and the
noise in the left-hand plot was at one-sixth the power.

a. Modify the program to generate two- and six-level signals.

b. Make a plot of the noise power versus the percentage of errors for two, four,
and six levels.
Exercise 14.14. Use noisychan.m to compare the power requirements for twolevel, four-level, and six-level transmissions. Fix the noise power at p=0.01, and
nd the error probability for four-level transmission. Experimentally nd the
power S that is required to make the two-level and six-level transmissions have
the same probability of error. Can you think of a way to calculate this?
Exercise 14.15. Consider the (asymmetric, nonuniformly spaced) alphabet consisting of the symbols 1, 1, 3, 4.
a. Find the amplitudes of this four-level signal with unity power.
b. Use noisychan.m to examine the noise performance of this transmission by
making a plot of the noise power versus the percentage of errors.
c. Compare this alphabet with 4-PAM using the standard alphabet 1, 3.
Which would you prefer?
There are two dierent problems that can keep a transmission system from
reaching the Shannon limit. The rst is that the source might not be coded with
maximum entropy, and this will be discussed next in Section 14.5. The second
is when dierent symbols experience dierent amounts of noise. Recall that the
plausibility argument for the channel capacity rested on the idea of the average

Coding

323

noise. When symbols encounter anything less than the average noise, then all is
well, since the average distance between levels is greater than the average noise.
But errors occur when symbols encounter more than the average amount of noise.
(This is why there are so many errors in the right-hand plot of Figure 14.4.) Good
coding schemes try to ensure that all symbols experience (roughly) the average
noise. This can be accomplished by grouping the symbols into clusters or blocks
that distribute the noise evenly among all the symbols in the block. Such error
coding is discussed in Section 14.6.

14.5

Source Coding
The results from Section 14.3 suggest that, all else being equal, it is preferable to
choose a code in which each symbol occurs with the same probability. But what if
the symbols occur with widely varying frequencies? Recall that this was shown
in Section 14.2 for English and other natural languages. There are two basic
approaches. The rst aggregates the letters into clusters, and provides a new
(longer) code word for each cluster. If properly chosen, the new code words can
occur with roughly the same probability. The second approach uses variablelength code words, assigning short codes to common letters like e and long
codes to infrequent letters like x. Perhaps the most common variable-length
code was that devised by Morse for telegraph operators, which used a sequence
of dots and dashes (along with silences of various lengths) to represent the
letters of the alphabet.
Before discussing how source codes can be constructed, consider an example
using the N = 4 code from Example 14.4(a) in which p(x1 ) = 0.5, p(x2 ) = 0.25,
and p(x3 ) = p(x4 ) = 0.125. As shown earlier, the entropy of this source is 1.75
bits/symbol, which means that there must be some way of coding the source so
that, on average, 1.75 bits are used for each symbol. The naive approach to this
source would use 2 bits for each symbol, perhaps assigning
x1 11, x2 10, x3 01, and x4 00.

(14.8)

An alternative representation is
x1 1, x2 01, x3 001, and x4 000,

(14.9)

where more-probable symbols use fewer bits, and less-probable symbols require
more. For instance, the string
x1 , x2 , x1 , x4 , x3 , x1 , x1 , x2
(in which each element appears with the expected frequency) is coded as
10110000011101.
This requires 14 bits to represent the eight symbols. The average is 14/8 = 1.75
bits per symbol, so this coding is as good as possible, since it equals the entropy.

324

Chapter 14. Coding

In contrast, the naive code of (14.8) requires 16 bits to represent the eight symbols for an average of 2 bits per symbol. One feature of the variable-length
code in (14.9) is that there is never any ambiguity about where it starts,
since any occurrence of a 1 corresponds to the end of a symbol. The naive
code requires knowledge of where the rst symbol begins. For example, the
string 01 10 11 00 1 is very dierent from 0 11 01 10 01 even
though they contain the same bits in the same order. Codes for which the
start and end are immediately recognizable are called instantaneous or prex
codes.
Since the entropy denes the smallest number of bits that can be used to
encode a source, it can be used to dene the eciency of a code
eciency =

entropy
.
average # of bits per symbol

(14.10)

Thus the eciency of the naive code (14.8) is 1.75/2 = 0.875 while the eciency of the variable-rate code (14.9) is 1. Shannons source coding theorem
says that, if an independent source has entropy H, then there exists a prex code in which the average number of bits per symbol is between H and
H + 1. Moreover, there is no uniquely decodable code that has smaller average length. Thus, if N symbols (each with entropy H) are compressed into
fewer than N H bits, information is lost, while information need not be lost
if N (H + 1) bits are used. Shannon has dened the goal towards which all
codes aspire, but provides no way of nding good codes for any particular
case.
Fortunately, D. A. Human proposed an organized procedure to build variablelength codes that are as ecient as possible. Given a set of symbols and their
probabilities, the procedure is as follows.
1. List the symbols in order of decreasing probability. These are the original
nodes.
2. Find the two nodes with the smallest probabilities, and combine them into
one new node, with probability equal to the sum of the two. Connect the new
nodes to the old ones with branches (lines).
3. Continue combining the pairs of nodes with the smallest probabilities. (If
there are ties, pick any of the tied symbols.)
4. Place a 0 or a 1 along each branch. The path from the rightmost node to the
original symbol denes a binary list, which is the code word for that symbol.
This procedure is probably easiest to understand by working through an example.
Consider again the rst N = 4 code from Example 14.4 in which the symbols have
probabilities p(x1 ) = 0.5, p(x2 ) = 0.25, and p(x3 ) = p(x4 ) = 0.125. Following the
foregoing procedure leads to the chart shown in Figure 14.5. In the rst step, x3
and x4 are combined to form a new node with probability equal to 0.25 (the sum
p(x3 ) + p(x4 )). Then this new node is combined with x2 to form a new node with
probability 0.5. Finally, this is combined with x1 to form the rightmost node.

Coding

Symbol

Probability

x1

0.5

x2
x3
x4

1
0
1
0

0.25

0.125

325

1
0

0.5

Figure 14.5 The Human code for the

source dened in the rst part of
Example 14.4 can be read directly from
this chart, which is constructed using
the procedure (1) to (4) in the text.

0.25

0.125

Each branch is now labeled. The convention used in Figure 14.5 is to place a 1
on the top and a 0 on the bottom (assigning the binary digits in another order
just relabels the code). The Human code for this source can be read from the
chart. Reading from the right-hand side, x1 corresponds to 1, x2 to 01, x3 to
001, and x4 to 000. This is the same code as in (14.9).
The Human procedure with a consistent branch-labeling convention always
leads to a prex code because all the symbols end the same (except for the
maximal-length symbol x4 ). More importantly, it always leads to a code that
has average length very near the optimal.
Exercise 14.16. Consider the source with N = 5 symbols with probabilities
p(x1 ) = 1/16, p(x2 ) = 1/8, p(x3 ) = 1/4, p(x4 ) = 1/16, and p(x5 ) = 1/2.
a. What is the entropy of this source?
b. Build the Human chart.
c. Show that the Human code is
x1 0001, x2 001, x3 01, x4 0000, and x5 1.
d. What is the eciency of this code?
e. If this source were encoded naively, how many bits per symbol would be
needed? What is the eciency?
Exercise 14.17. Consider the source with N = 4 symbols with probabilities
p(x1 ) = 0.3, p(x2 ) = 0.3, p(x3 ) = 0.2, and p(x4 ) = 0.2.
a.
b.
c.
d.

What is the entropy of this source?

Build the Human code.
What is the eciency of this code?
If this source were encoded naively, how many bits per symbol would be
needed? What is the eciency?

Exercise 14.18. Build the Human chart for the source dened by the 26
English letters (plus space) and their frequency in the Wizard of Oz as given
in (14.4).

326

Chapter 14. Coding

The Matlab program codex.m demonstrates how a variable-length code can

be encoded and decoded. The rst step generates a 4-PAM sequence with the
probabilities used in the rst part of Example 14.4. In the code, the symbols
are assigned numerical values {1, 3}. The symbols, their probabilities, the
numerical values, and the variable-length Human code are as follows.
symbol probability value Human code
0.5
+1
1
x1
0.25
1
01
x2
x3
0.125
+3
001
0.125
3
000
x4
This Human code was derived in Figure 14.5. For a length-m input sequence, the
second step replaces each symbol value with the appropriate binary sequence,
and places the output in the vector cx.
Listing 14.6. codex.m step 2: encode the sequence using Human code
j =1;
cx = [ ] ;
f o r i =1:m
i f x ( i )==+1,
e l s e i f x ( i )==1,
e l s e i f x ( i )==+3,
e l s e i f x ( i )==3,
end
end

cx ( j
cx ( j
cx ( j
cx ( j

)=[1];
: j +1)=[0 ,1];
: j +2)=[0 ,0 ,1];
: j +2)=[0 ,0 ,0];

j=j +1;
j=j +2;
j=j +3;
j=j +3;

The third step carries out the decoding. Assuming the encoding and decoding
have been done properly, cx is transformed into the output y, which should be
the same as the original sequence x.
Listing 14.7. codex.m step 3: decode the variable-length sequence
j =1; i =1;
y=[];
while i<=length ( cx )
i f cx ( i : i ) = = [ 1 ] , . . .
y ( j )=+1; i=i +1; j=j +1;
e l s e i f cx ( i : i + 1 ) = = [ 0 , 1 ] , . . .
y ( j )= 1; i=i +2; j=j +1;
e l s e i f cx ( i : i + 2 ) = = [ 0 , 0 , 1 ] , . . .
y ( j )=+3; i=i +3; j=j +1;
e l s e i f cx ( i : i + 2 ) = = [ 0 , 0 , 0 ] , . . .
y ( j )= 3; i=i +3; j=j +1; end
end

Indeed, running the program codex.m (which contains all three steps) gives a
perfect decoding.
Exercise 14.19. Mimicking the code in codex.m, create a Human encoder and
decoder for the source dened in Example 14.16.
Exercise 14.20. Use codex.m to investigate what happens when the probabilities of the source alphabet change.

Coding

327

a. Modify step 1 of the program so that the elements of the input sequence have
probabilities
x1 0.1, x2 0.1, x3 0.1, and x4 0.7.

(14.11)

b. Without changing the Human encoding to account for these changed probabilities, compare the average length of the coded data vector cx with the
average length of the naive encoder (14.8). Which does a better job compressing the data?
c. Modify the program so that the elements of the input sequence all have the
same probability, and answer the same question.
d. Build the Human chart for the probabilities dened in (14.11).
e. Implement this new Human code and compare the average length of the
coded data cx with the previous results. Which does a better job compressing
the data?
Exercise 14.21. Using codex.m, implement the Human code from Exercise 14.18. What is the length of the resulting data when applied to the text
of the Wizard of Oz? What rate of data compression has been achieved?
Source coding is used to reduce the redundancy in the original data. If the
letters in the Wizard of Oz were independent, then the Human coding in Exercise 14.21 would be optimal: no other coding method could achieve a better compression ratio. But the letters are not independent. More sophisticated schemes
would consider not just the raw probabilities of the letters, but the probabilities
of pairs of letters, or of triplets, or more. As suggested by the redundancy studies in Section 14.2, there is a lot that can be gained by exploiting higher-order
relationships between the symbols.
Exercise 14.22. Zipped les (usually with a .zip extension) are a popular
form of data compression for text (and other data) on the web. Download a
handful of .zip les. Note the le size when the le is in its compressed form and
the le size after decompressing (unzipping) the le. How does this compare
with the compression ratio achieved in Exercise 14.21?
Exercise 14.23. Using the routine writetext.m (this le, which can be found
on the website, uses the Matlab command fwrite), write the Wizard of Oz
text to a le OZ.doc. Use a compression routine (uuencode on a Unix or Linux
machine, zip on a Windows machine, or stut on a Mac) to compress OZ.doc.
Note the le size when the le is in its compressed form, and the le size after
decompressing. How does this compare with the compression ratio achieved in
Exercise 14.21?

328

Chapter 14. Coding

14.6

Channel Coding
The job of channel or error-correcting codes is to add some redundancy to a
signal before it is transmitted so that it becomes possible to detect when errors
have occurred and to correct them, when possible.
Perhaps the simplest technique is to send each bit three times. Thus, in order
to transmit a 0, the sequence 000 is sent. In order to transmit a 1, 111 is sent.
This is the encoder. At the receiver, there must be a decoder. There are eight
possible sequences that can be received, and a majority-rules decoder assigns
000 0
101 1

001 0
110 1

010 0
011 1

100 0
111 1.

(14.12)

This encoder/decoder can identify and correct any isolated single error and so
the transmission has smaller probability of error. For instance, assuming no more
than one error per block, if 101 was received, then the error must have occurred
in the middle bit, whereas if 110 was received, then the error must have been
in the third bit. But the majority-rules coding scheme is costly: three times the
number of symbols must be transmitted, which reduces the bit rate by a factor of
three. Over the years, many alternative schemes have been designed to reduce the
probability of error in the transmission without incurring such a heavy penalty.
Linear block codes are popular because they are easy to design, easy to implement, and have a number of useful properties. With n > k, an (n, k) linear
code operates on sets of k symbols, and transmits a length-n code word for each
set. Each code is dened by two matrices: the k by n generator matrix G, and
the n k by n parity-check matrix H. In outline, the operation of the code is
as follows.
1. Collect k symbols into a vector x = {x1 , x2 , . . . , xk }.
2. Transmit the length-n code word c = xG.
3. At the receiver, the vector y is received. Calculate yH T .
4. If yH T = 0, then no errors have occurred.
5. When yH T = 0, errors have occurred. Look up yH T in a table of syndromes, which contains a list of all possible received values and the most
likely code word to have been transmitted, given the error that occurred.
6. Translate the corrected code word back into the vector x.
The simplest way to understand this is to work through an example in detail.

14.6.1

A (5, 2) Binary Linear Block Code

To be explicit, consider the case of a (5, 2) binary code with generator matrix


10101
G=
(14.13)
01011

Coding

329

Table 14.1. Modulo 2 arithmetic

+
0
1

0
0
1

0
1

1
1
0

0
0
0

1
0
1

and parity-check matrix

1
0

HT =
1
0
0

01
1 1

0 0
.
1 0

(14.14)

01

This code bundles the bits into pairs, and the four corresponding code words are
x1 = 00 c1 = x1 G = 00000,
x2 = 01 c2 = x2 G = 01011,
x3 = 10 c3 = x3 G = 10101,
and
x4 = 11 c4 = x4 G = 11110.
There is one subtlety. The arithmetic used in the calculation of the code words
(and indeed throughout the linear-block-code method) is not standard. Because
the input source is binary, the arithmetic is also binary. Binary addition and
multiplication are shown in Table 14.1. The operations of binary arithmetic may
be more familiar as exclusive OR (binary addition) and logical AND (binary
multiplication).
In eect, at the end of every calculation, the answer is taken modulo 2. For
instance, in standard arithmetic, x4 G = 11112. The correct code word c4 is
found by reducing each calculation modulo 2. In Matlab, this is done with
mod(x4*g,2), where x4=[1,1] and g is dened as in (14.13). In modulo 2 arithmetic, 1 represents any odd number and 0 represents any even number. This is
also true for negative numbers so that, for instance, 1 = +1 and 4 = 0.
After transmission, the received signal y is multiplied by H T . If there were no
errors in transmission, then y is equal to one of the four code words ci . With H
dened as in (14.14), c1 H T = c2 H T = c3 H T = c4 H T = 0, where the arithmetic
is binary, and where 0 means the zero vector of size 1 by 3 (in general, 1 by
(n k)). Thus yH T = 0 and the received signal is one of the code words.
However, when there are errors, yH T = 0, and the value can be used to determine the most likely error to have occurred. To see how this works, rewrite
y = c + (y c) c + e,

330

Chapter 14. Coding

Table 14.2. Syndrome table for the binary (5, 2) code with generator matrix (14.13) and
parity-check matrix (14.14)

Syndrome eH T
000
001
010
011
100
101
110
111

Most likely error e

00000
00001
00010
01000
00100
10000
11000
10010

where e represents the error(s) that have occurred in the transmission. Note that
yH T = (c + e)H T = cH T + eH T = eH T ,
since cH T = 0. The value of eH T is used by looking in the syndrome table,
Table 14.2. For example, suppose that the symbol x2 = 01 is transmitted using
the code c2 = 01011. But an error occurs in transmission so that y = 11011 is
received. Multiplication by the parity-check matrix gives yH T = eH T = 101.
Looking this up in the syndrome table shows that the most likely error was
10000. Accordingly, the most likely code word to have been transmitted was
y e = 11011 10000 = 01011, which is indeed the correct code word c2 .
On the other hand, if more than one error occurred in a single symbol, then the
(5, 2) code cannot necessarily nd the correct code word. For example, suppose
that the symbol x2 = 01 is transmitted using the code c2 = 01011 but that two
errors occur in transmission so that y = 00111 is received. Multiplication by the
parity-check matrix gives yH T = eH T = 111. Looking this up in the syndrome
table shows that the most likely error was 10010. Accordingly, the most likely
symbol to have been transmitted was y e = 00111 + 10010 = 10101, which is
the code word c3 corresponding to the symbol x3 , and not c2 .
The syndrome table can be built as follows. First, take each possible single
error pattern, that is, each of the n = 5 es with exactly one 1, and calculate eH T
for each. As long as the columns of H are nonzero and distinct, each error pattern
corresponds to a dierent syndrome. To ll out the remainder of the table, take
each of the possible double errors (each of the es with exactly two 1s) and
calculate eH T . Pick two that correspond to the remaining unused syndromes.
Since there are many more possible double errors n(n 1) = 20 than there are
syndromes (2nk = 8), these are beyond the ability of the code to correct.
The Matlab program blockcode52.m shows details of how this encoding and
decoding proceeds. The rst part denes the relevant parameters of the (5, 2)
binary linear block code: the generator g, the parity-check matrix h, and the

Coding

331

syndrome table syn. The rows of syn are ordered so that the binary digits of
eH T can be used to directly index into the table.
Listing 14.8. blockcode52.m part 1: denition of (5, 2) binary linear block code, the
generator and parity-check matrices
g =[1 0 1 0 1 ;
0 1 0 1 1];
h=[1 0 1 0 0 ;
0 1 0 1 0;
1 1 0 0 1];
% th e f o u r code words cw=x * g (mod 2 )
x ( 1 , : ) = [ 0 0 ] ; cw ( 1 , : ) =mod ( x ( 1 , : ) * g , 2 ) ;
x ( 2 , : ) = [ 0 1 ] ; cw ( 2 , : ) =mod ( x ( 2 , : ) * g , 2 ) ;
x ( 3 , : ) = [ 1 0 ] ; cw ( 3 , : ) =mod ( x ( 3 , : ) * g , 2 ) ;
x ( 4 , : ) = [ 1 1 ] ; cw ( 4 , : ) =mod ( x ( 4 , : ) * g , 2 ) ;
% th e syndrome t a b l e
syn =[0 0 0 0 0 ;
0 0 0 0 1;
0 0 0 1 0;
0 1 0 0 0;
0 0 1 0 0;
1 0 0 0 0;
1 1 0 0 0;
1 0 0 1 0];

The second part carries out the encoding and decoding. The variable p species
the chance that bit errors will occur in the transmission. The code words c are
constructed using the generator matrix. The received signal is multiplied by the
parity-check matrix h to give the syndrome, which is then used as an index into
the syndrome table (matrix) syn. The resulting most likely error is subtracted
from the received signal, and this is the corrected code word that is translated
back into the message. Because the code is linear, code words can be translated
back into the message using an inverse matrix9 and there is no need to store
all the code words. This becomes important when there are millions of possible
code words, but when there are only four it is not crucial. The translation is
done in blockcode52.m in the for j loop by searching.
Listing 14.9. blockcode52.m part 2: encoding and decoding data
p=.1;
m=10000;
dat = 0 . 5 * ( sign ( rand ( 1 ,m) 0 . 5 ) + 1 ) ;
f o r i = 1 : 2 :m
c=mod ( [ dat ( i ) dat ( i + 1 )] * g , 2 ) ;
f o r j =1: length ( c )
i f rand<p , c ( j )=c ( j )+1; end
end
y=c ;
eh=mod ( y * h , 2 ) ;
e h i n d=eh (1) * 4+ eh (2) * 2+ eh ( 3 ) + 1 ;
e=syn ( ehind , : ) ;
y=mod ( ye , 2 ) ;
9

% pr obabi l i ty of bit f l i p
% l e n g t h o f message
% m random 0 s and 1 s
% b u i l d codeword
% flip
%
%
%
%
%

b i t s with prob p

received signal
m u l t i p l y by p a r i t y c h e c k h
tu r n syndrome i n t o i n d e x
e r r o r from syndrome t a b l e
add e to c o r r e c t e r r o r s

This is explored in the context of blockcode52.m in Exercise 14.26.

332

Chapter 14. Coding

f o r j =1:max( s i z e ( x ) )
% g e t message from codewords
i f y==cw ( j , : ) , z ( i : i +1)=x ( j , : ) ; end
end
end
e r r=sum( abs ( zdat ) )
% how many e r r o r s o c c u r r e d

Running blockcode52.m with the default parameters of 10% bit errors and
length m=10000 will give about 400 errors, a rate of about 4%. Actually, as
will be shown in the next section, the performance of this code is slightly better
than these numbers suggest, because it is also capable of detecting certain errors
that it cannot correct, and this feature is not implemented in blockcode52.m.
Exercise 14.24. Use blockcode52.m to investigate the performance of the
binary (5, 2) code. Let p take on a variety of values p = 0.001, 0.01, 0.02, 0.05,
0.1, 0.2, and 0.5 and plot the percentage of errors as a function of the percentage
of bits ipped.
Exercise 14.25. This exercise compares the performance of the (5, 2) block
code in a more realistic setting and provides a good warm-up exercise for the
receiver to be built in Chapter 15. The program nocode52.m (all Matlab les are
available on the website) provides a template with which you can add the block
coding into a real transmitter and receiver pair. Observe, in particular, that
the block coding is placed after the translation of the text into binary but before
the translation into 4-PAM (for transmission). For eciency, the text is encoded
using text2bin.m (recall Example 8.2). At the receiver, the process is reversed:
the raw 4-PAM data are translated into binary, then decoded using the (5, 2)
block decoder, and nally translated back into text (using bin2text.m) where
you can read it. Your task in this problem is to experimentally verify the gains
possible when using the (5, 2) code. First, merge the programs blockcode52.m
and nocode52.m. Measure the number of errors that occur as noise is increased
(the variable varnoise scales the noise). Make a plot of the number of errors
as the variance increases. Compare this with the number of errors that occur as
the variance increases when no coding is used (i.e., running nocode52.m without
modication).
Exercise 14.26. Use the matrix ginv=[1 1;1 0 ;0 0;1 0;0 1]; to replace
the for j loop in blockcode52.m. Observe that this reverses the eect of constructing the code words from the x since cw*ginv=x (mod 2).
Exercise 14.27. Implement the simple majority-rules code described in (14.12).
a. Plot the percentage of errors after coding as a function of the number of
symbol errors.
b. Compare the performance of the majority-rules code to the (5, 2) block code.
c. Compare the data rate required by the majority-rules code with that required
by the (5, 2) code and with the naive (no coding) case.

Coding

14.6.2

333

Minimum Distance of a Linear Code

In general, linear codes work much like the example in the previous section,
although the generator matrix, parity-check matrix, and syndrome table are
unique to each code. The details of the arithmetic may also be dierent when
the code is not binary. Two examples will be given later. This section discusses
the general performance of linear block codes in terms of the minimum distance
of a code, which species how many errors the code can detect and how many
errors it can correct.
A code C is a collection of code words ci , which are n-vectors with elements
drawn from the same alphabet as the source. An encoder is a rule that assigns
a k-length message to each codeword.

Example 14.5. The code words of the (5, 2) binary code are 00000, 01011, 10101,
and 11110, which are assigned to the four input pairs 00, 01, 10, and 11, respectively. The Hamming distance10 between any two elements in C is equal to the
number of places in which they disagree. For instance, the distance between 00000
and 01011 is 3, which is written d(00000, 01011) = 3. The distance between 1001
and 1011 is d(1001, 1011) = 1. The minimum distance of a code C is the smallest
distance between any two code words. In symbols,
dmin = min d(ci , cj ),
i=j

where ci C.
Exercise 14.28. Show that the minimum distance of the (5, 2) binary linear
block code is dmin = 3.
Exercise 14.29. Write down all code words for the majority-rules code (14.12).
What is the minimum distance of this code?
Exercise 14.30. A code C has four elements {0000, 0101, 1010, 1111}. What
is the minimum distance of this code?
Let Di (t) be the decoding sets of all possible received signals that are less
than t away from ci . For instance, the majority-rules code has two code words,
and hence two decoding sets. With t = 1, these are
D1 (1) = {000, 001, 100, 010},
D2 (1) = {111, 110, 011, 101}.
10

(14.15)

Named after R. Hamming, who also created the Hamming blip as a windowing function.
Software Receiver Design adopted the blip in previous chapters as a convenient pulse
shape.

334

Chapter 14. Coding

When any of the elements in D1 (1) are received, the code word c1 = 0 is used;
when any of the elements in D2 (1) are received, the code word c2 = 1 is used.
For t = 0, the decoding sets are
D1 (0) = {000} and D2 (0) = {111}.

(14.16)

In this case, when 000 is received, c1 is used; when 111 is received, c2 is used.
When the received bits are in neither of the Di , an error is detected, though it
cannot be corrected. When t > 1, the Di (t) are not disjoint and hence cannot be
used for decoding.
Exercise 14.31. What are the t = 0 decoding sets for the four-element code in
Exercise 14.30? Are the t = 1 decoding sets disjoint?
Exercise 14.32. Write down all possible disjoint decoding sets for the (5, 2)
linear binary block code.
One use of decoding sets lies in their relationship with dmin . If 2t < dmin , then
the decoding sets are disjoint. Suppose that the code word ci is transmitted over
a channel, but that c (which is obtained by changing at most t components of ci )
is received. Then c still belongs to the correct decoding set Di , and is correctly
decoded. This is an error-correction code that handles up to t errors.
Now suppose that the decoding sets are disjoint with 2t + s < dmin , but that
t < d(c, ci ) t + s. Then c is not a member of any decoding set. Such an error
cannot be corrected by the code, though it is detected. The following example
shows how the ability to detect errors and the ability to correct them can be
traded o.

Example 14.6. Consider again the majority-rules code C with two elements
{000, 111}. This code has dmin = 3 and can be used as follows.
1. t = 1, s = 0. In this mode, using decoding sets (14.15), code words could suer
any single error and still be correctly decoded. But, if two errors occurred,
the message would be incorrect.
2. t = 0, s = 2. In this mode, using decoding sets (14.16), the code word could
suer up to two errors and the error would be detected, but there would be
no way to correct it with certainty.

Example 14.7. Consider the code C with two elements {0000000, 1111111}.
Then, dmin = 7. This code can be used in the following ways.
1. t = 3, s = 0. In this mode, the code word could suer up to three errors and
still be correctly decoded. But, if four errors occurred, the message would be
incorrect.

Coding

335

Table 14.3. Syndrome table for the binary (7, 3) code

Syndrome
eH T
0000
0001
0010
0100
1000
1101
1011
0111
0011
0110
1100
0101
1010
1001
1110
1111

Most likely
error e
0000000
0000001
0000010
0000100
0001000
0010000
0100000
1000000
0000011
0000110
0001100
0011000
0001010
0010100
0111000
0010010

2. t = 2, s = 2. In this mode, if the codeword suered up to two errors, then it

would be correctly decoded. If there were three or four errors, then the errors
would be detected, but, because they cannot be corrected with certainty, no
(incorrect) message is generated.
Thus, the minimum distance of a code is a resource that can be allocated
between error detection and error correction. How to trade these o is a system
design issue. In some cases, the receiver can ask for a symbol to be retransmitted
when an error occurs (for instance, in a computer modem or when reading a le
from disk), and it may be sensible to allocate dmin to detecting errors. In other
cases (such as broadcasting), it is more common to focus on error correction.
The discussion in this section so far is completely general; that is, the denition
and results on minimum distance apply to any code of any size, whether linear
or nonlinear. There are two problems with large nonlinear codes.

r It is hard to specify codes with large dmin .

r Implementing coding and decoding can be expensive in terms of memory and
computational power.
To emphasize this, consider a code that combines binary digits into clusters of
56 and codes these clusters using 64 bits. Such a code requires about 10101 code
words. Considering that the estimated number of elementary particles in the
universe is about 1080 , this is a problem. When the code is linear, however, it is
not necessary to store all the code words; they can be generated as needed. This

336

Chapter 14. Coding

was noted in the discussion of the (5, 2) code of the previous section. Moreover,
nding the minimum distance of a linear code is also easy, since dmin is equal
to the smallest number of nonzero coordinates in any code word (not counting
the zero code word). Thus dmin can be calculated directly from the denition
by nding the distances between all the code words, or by nding the code word
that has the smallest number of 1s. For instance, in the (5, 2) code, the two
elements 01011 and 10101 each have exactly three nonzero terms.

14.6.3

Some More Codes

This section gives two examples of (n, k) linear codes. If the generator matrix G
has the form
G = [Ik |P ],

(14.17)

where Ik is the k by k identity matrix and P is some k by n k matrix, then

P
[Ik |P ] = 0,
(14.18)
Ink
where the 0 is the k by n k matrix of all zeros. Hence, dene H = [P |Ink ].
Observe that the (5, 2) code is of this form, since in binary arithmetic 1 = +1
and so P = P .

Example 14.8. A (7, 3) binary code has generator matrix

1000111
G = 0 1 0 1 0 1 1
0011101
and parity-check matrix

0 1
1 0

1 1

T
H =
1 0

0 1

0 0
0 0

1 1
1 1

0 1

.
0 0

0 0

1 0
0 1

The syndrome Table 14.3 is built by calculating which error pattern is most likely
(i.e., has the fewest bits ipped) for each given syndrome eH T . This code has
dmin = 4, and hence the code can correct any 1-bit errors, 7 (out of 21) possible
2-bit errors, and 1 of the many 3-bit errors.

Coding

337

Table 14.4. Modulo 5 arithmetic

+
0
1
2
3
4

0
0
1
2
3
4

1
1
2
3
4
0

2
2
3
4
0
1

3
3
4
0
1
2

4
4
0
1
2
3

0
1
2
3
4

0
0
0
0
0
0

1
0
1
2
3
4

2
0
2
4
1
3

3
0
3
1
4
2

4
0
4
3
2
1

Exercise 14.33. Using the code from blockcode52.m, implement the binary
(7, 3) linear block code. Compare its performance and eciency with the (5, 2)
code and the majority-rules code.
a. For each code, plot the percentage p of bit ips in the channel versus the
percentage of bit ips in the decoded output.
b. For each code, what is the average number of bits transmitted for each bit in
the message?
Sometimes, when the source alphabet is not binary, the elements of the code
words are also not binary. In this case, using the binary arithmetic of Table 14.1 is
inappropriate. For example, consider a source alphabet with ve symbols labeled
0, 1, 2, 3, 4. Arithmetic operations for these elements are addition and multiplication modulo 5, which are dened in Table 14.4. These can be implemented
in Matlab using the mod function. For some source alphabets, the appropriate
arithmetic operations are not modulo operations, and in these cases, it is normal
to simply dene the desired operations via tables like Tables 14.1 and 14.4.

Example 14.9. A (6, 4) code using a q = 5-element source alphabet has generator
matrix

100044
0 1 0 0 4 3

G=
0 0 1 0 4 2
000141
and parity-check matrix

1
4 4
4 3 1

4 2 1
T
H =
=
4 1 1

1 0 1
0
0 1

1
2

3
,
4

0
1

338

Chapter 14. Coding

Table 14.5. Syndrome table for the q = 5 source alphabet (6, 4) code

Syndrome
eH T
00
01
10
14
13
12
11
02
20
23
21
24
22
03
30
32
34
31
33
04
40
41
42
43
44

Most likely
error e
000000
000001
000010
000100
001000
010000
100000
000002
000020
000200
002000
020000
200000
000003
000030
000300
003000
030000
300000
000004
000040
000400
004000
040000
400000

since, in modulo 5 arithmetic, 4 = 1, 3 + 2, 2 = 3, and 1 = 4. Observe

that these t in the general form of (14.17) and (14.18). Table 14.5 lists the
qnk = 564 = 25 syndromes and corresponding errors.
The code in this example corrects all one-symbol errors (and no others).
Exercise 14.34. Find all the code words in the q = 5 (6, 4) linear block code
from Example 14.9.
Exercise 14.35. What is the minimum distance of the q = 5 (6, 4) linear block
code from Example 14.9?
Exercise 14.36. Mimicking the code in blockcode52.m, implement the q = 5
(6, 4) linear block code from Example 14.9. Compare its performance with the
(5, 2) and (7, 3) binary codes in terms of

Coding

339

a. performance in correcting errors,

b. data rate.
Be careful: how can a q = 5 source alphabet be compared fairly with a binary
alphabet? Should the comparison be in terms of percentage of bit errors or
percentage of symbol errors?

14.7

Encoding a Compact Disc

The process of writing to and reading from a compact disc is involved. The
essential idea in optical media is that a laser beam bounces o the surface of
the disc. If there is a pit, then the light travels a bit further than if there is no
pit. The distances are controlled so that the extra time required by the round
trip corresponds to a phase shift of 180. Thus, the light traveling back interferes
destructively if there is a pit, whereas it reinforces constructively if there is no
pit. The strength of the beam is monitored to detect a 0 (a pit) or a 1 (no pit).
While the complete system can be made remarkably accurate, the reading and
writing procedures are prone to errors. This is a perfect application for errorcorrecting codes! The encoding procedure is outlined in Figure 14.6. The original
signal is digitized at 44,100 samples per second in each of two stereo channels.
Each sample is 16 bits, and the eective data rate is 1.41 Mbps (megabits per
second). The Cross Interleave ReedSolomon Code (CIRC) encoder (described
shortly) has an eective rate of about 3/4, and its output is at 1.88 Mbps. Then
control and timing information is added, which contains the track and subtrack
numbers that allow CD tracks to be accessed rapidly. The EFM (Eight-toFourteen Module) is an encoder that spreads the audio information in time by
changing each possible 8-bit sequence into a predened 14-bit sequence so that
each one is separated by at least two (and at most ten) zeros. This is used to
help the tracking mechanism. Reading errors on a CD often occur in clusters (a
small scratch may be many hundreds of bits wide) and interleaving distributes
the errors so that they can be corrected more eectively. Finally, a large number
of synchronization bits is added. These are used by the control mechanism of
the laser to tell it where to shine the beam in order to nd the next bits. The
nal encoded data rate is 4.32 Mbps. Thus, about 1/3 of the bits on the CD are
actual data, and about 2/3 of the bits are present to help the system function
and to detect (and/or correct) errors when they occur.
The CIRC encoder consists of two special linear block codes called Reed
Solomon codes (which are named after their inventors). Both use q = 256 (8bit) symbols, and each 16-bit audio sample is split into two code words. The
rst code is a (32,28) linear code with dmin = 5, and the second code is a linear
(28,24) code, also with dmin = 5. These are nonbinary and use special arithmetic
operations dened by the Galois eld with 256 symbols. The encoding is split

340

Chapter 14. Coding

Music

A/D

44.1 kHz per channel

at 16 bits/sample

CIRC
Encoder

ReedSolomon
Linear Block Codes

MUX

"Control and timing

track/subtracks"

EFM

"Eight-to-Fourteen Module"

MUX

"Synchronization"

1.41 Mbps

1.88 Mbps

1.92 Mbps

Figure 14.6 CDs can be used

for audio or for data. The
encoding procedure is the
same, though decoding may
be done dierently for
dierent applications.

4.32 Mbps
Optical media

into two separate codes so that an interleaver can be used between them. This
spreads out the information over a larger range and helps to spread out the errors
(making them easier to detect and/or correct).
The encoding process on the CD is completely specied, but each manufacturer
can implement the decoding as they wish. Accordingly, there are many choices.
For instance, the ReedSolomon codes can be used to correct two errors each,
or to detect up to ve errors. When errors are detected, a common strategy is
to interpolate the audio, which may be transparent to the listener as long as the
error rate is not too high. Manufacturers may also choose to mute the audio when
the error rate is too high. For data purposes, the controller can also ask that the
data be reread. This may allow correction of the error when it was caused by
mistracking or some other transitory phenomenon, but will not be eective if the
cause is a defect in the medium.

For Further Reading

The paper that started information theory is still a good read half a century
after its initial publication:

r C. E. Shannon, A Mathematical Theory of Communication, The Bell System Technical Journal, vol. 27, pp. 379423 and 623656, July and October,
1948.

Step 5: Putting It All Together

The projects of Chapters 15 and 16 integrate all the xes of the fourth step
into the receiver structure of the third step to create fully functional digital
receivers. The well-fabricated receiver is robust with respect to distortions such
as those caused by noise, multipath interference, timing inaccuracies, and clock
mismatches.
After building the components, testing them, assembling them into a receiver,
and testing the full design, your receiver is ready. Congratulations. You have
earned the degree of Master of Digital Radio. You are now ready to conquer the
world!

Master

of
l
Digita

Radio

____
_____
_____ me here
a
n
your

15 Make It So

This chapter describes a software-dened-radio design project called M6 , the Mix

n Match Mostly Marvelous Message Machine. The M6 transmission standard
is specied so that the receiver can be designed using the building blocks of the
preceding chapters. The DSP portion of the M6 can be simulated in Matlab
by combining the functions and subroutines from the examples and exercises of
the previous chapters.
The input to the digital portion of the M6 receiver is a sampled signal at
intermediate frequency (IF) that contains several simultaneous messages, each
transmitted in its own frequency band. The original message is text that has
been converted into symbols drawn from a 4-PAM constellation, and the pulse
shape is a square-root raised cosine. The sample frequency can be less than twice
the highest frequency in the analog IF signal, but it must be suciently greater
than the inverse of the transmitted symbol period to be twice the bandwidth of
the baseband signal. The successful M6 Matlab program will demodulate, synchronize, equalize, and decode the signal, so it is a fully operational softwaredened receiver (although it is not intended to work in real time). The receiver
must overcome multiple impairments. There may be phase noise in the transmitter oscillator. There may be an oset between the frequency of the oscillator
in the transmitter and the frequency of the oscillator in the receiver. The pulse
clocks in the transmitter and receiver may dier. The transmission channel may
be noisy. Other users in spectrally adjacent bands may be actively transmitting
at the same time. There may be intersymbol interference caused by multipath
channels.
The next section describes the transmitter, the channel, and the analog front
end of the receiver. Then Section 15.2 makes several generic observations about
receiver design, and proposes a methodology for the digital receiver design. The
nal section describes the receiver design challenge that serves as the culminating
design experience of this book. Actually building the M6 receiver, however, is
left to you. You will know that your receiver works when you can recover the
mystery message hidden inside the received signal.

343

Make It So

15.1

How the Received Signal Is Constructed

Receivers cannot be designed in a vacuum; they must work in tandem with a particular transmitter. Sometimes, a communication system designer gets to design
both ends of the system. More often, however, the designer works on one end or
the other, with the goal of making the signal in the middle meet some standard
specications. The standard for the M6 is established on the transmitted signal,
and consists, in part, of specications on the allowable bandwidth and on the
precision of its carrier frequency. The standard also species the source constellation, the modulation, and the coding schemes to be used. The front end of the
receiver provides some bandpass ltering, downconversion to IF, and automatic
gain control prior to the sampler.
This section describes the construction of the sampled IF signal that must
be processed by the M6 receiver. The system that generates the analog received
signal is shown in block-diagram form in Figure 15.1. The front end of the receiver
that turns this into a sampled IF signal is shown in Figure 15.2.
The original message in Figure 15.1 is a character string of English text. Each
character is converted into a 7-bit binary string according to the ASCII conversion format (e.g., the letter a is 1100001 and the letter M is 1001101), as in
Example 8.2. The bit string is coded using the (5, 2) linear block code specied in
blockcode52.m which associates a 5-bit code with each pair of bits. The output
of the block code is then partitioned into pairs that are associated with the four

Text
message Character Bits
Coding
to binary
conversion

Symbols
Scaling
si
Trigger

Baseband
signal
Pulse
shape

Transmitted
passband
signal

Modulation

Adjacent
users

Channel

Broadband
noise

Analog
received
signal

includes periodic i Tt + t
marker and training
sequence

includes phase noise

Figure 15.1 Construction of the received signal.

Analog
received
signal
Bandpass
filter

Downconversion
to IF

Automatic
gain
control

Sampled
received
signal

Figure 15.2 Front end of

the receiver.
kTs

344

Chapter 15. Make It So

integers of a 4-PAM alphabet 1 and 3 via the mapping

11
10
01
00

+3
+1
1
3

(15.1)

as in Example 8.1. Thus, if there are n letters, there are 7n (uncoded) bits, 7n( 52 )
coded bits, and 7n( 25 )( 12 ) 4-PAM symbols. These mappings are familiar from Section 8.1, and are easy to use with the help of the Matlab functions bin2text.m
and text2bin.m. Exercise 14.25 provides several hints to help implement the M6
encoding, and the Matlab function nocode52.m outlines the necessary transformations from the original text into a sequence of 4-PAM symbols s[i].
In order to decode the message at the receiver, the recovered symbols must
be properly grouped and the start of each group must be located. To aid this
frame synchronization, a marker sequence is inserted into the symbol stream at
the start of every block of 100 letters (at the start of every 875 symbols). The
header/training sequence that starts each frame is given by the phrase
A0Oh well whatever Nevermind

(15.2)

which codes into 245 4-PAM symbols and is assumed to be known at the receiver.
This marker text string can be used as a training sequence by the adaptive
equalizer. The unknown message begins immediately after each training segment.
Thus, the M6 symbol stream is a coded message periodically interrupted by the
same marker/training clump.
As indicated in Figure 15.1, pulses are initiated at intervals of Tt seconds,
and each is scaled by the 4-PAM symbol value. This translates the discretetime symbol sequence s[i] (composed of the coded message interleaved with the
marker/training segments) into a continuous-time signal

s[i](t iTt t ).
s(t) =
i

The actual transmitter symbol period Tt is required to be within 0.01% of a

nominal M6 symbol period T = 6.4 s. The transmitter symbol period clock
is assumed to be steady enough that the timing oset t and its period Tt are
eectively time-invariant over the duration of a single frame.
Details of the M6 transmission specications are given in Table 15.1. The
pulse-shaping lter P (f ) is a square-root raised cosine lter symmetrically truncated to eight symbol periods. The rollo factor of the pulse-shaping lter is
xed within some range and is known at the receiver, though it could take on
dierent values with dierent transmissions. The (half-power) bandwidth of the
square-root raised cosine pulse could be as large as 102 kHz for the nominal T .
With double-sideband modulation, the pulse-shape bandwidth doubles so that
each passband FDM signal will need a bandwidth at least 204 kHz wide.

Make It So

345

Table 15.1. M6 Transmitter specications

Symbol source alphabet

Assigned intermediate frequency
Nominal symbol period
SRRC pulse-shape rollo factor
FDM user slot allotment
Width of SRRC pulse shape
Frame marker/training sequence
Frame marker sequence period
Time-varying IF carrier phase
IF oset
Timing oset
Symbol period oset
Intersymbol interference
Sampler frequency

1, 3
2 MHz
6.4 s
[0.1, 0.3]
204 kHz
8 clock periods
See (15.2)
1120 symbols
Filtered white noise
Fixed, < 0.01%
Fixed
Fixed, < 0.01%
Max. delay spread = 7 symbols
850 kHz

The channel may be near ideal (i.e., a unit gain multisymbol delay) or it may
have signicant intersymbol interference. In either case, the impulse response of
the channel is unknown at the receiver, though an upper bound on its delay
spread may be available. There are also disturbances that may occur during the
transmission. These may be wideband noise with at power spectral density or
they may be narrowband interferers, or both. They are unknown at the receiver.
The achieved intermediate frequency is required to be within 0.01% of its
assigned value. The carrier phase (t) is unknown to the receiver and may vary
over time, albeit slowly. This means that the phase of the intermediate-frequency
signal presented to the receiver sampler may also vary.
The bandpass lter before the downconverter in the front end of the receiver
in Figure 15.2 partially attenuates adjacent 204-kHz-wide FDM user bands. The
automatic gain control is presumed locked and xed over each transmission.
The free-running sampler frequency of 850 kHz is well above twice the 102 kHz
baseband bandwidth of the user of interest. This is necessary for the baseband
analog signal interpolator used in the timer in the DSP portion of the receiver
in Figure 15.3. However, the sampler frequency is not above twice the highest
frequency of the IF signal. This means that the sampled received signal has
replicated the spectrum at the output of the front end analog downconverter
lowpass lter to frequencies between zero and IF.

15.2

A Design Methodology for the M6 Receiver

Before describing the specic design requirements that must be met by a successful M6 receiver, this section makes some generic remarks about a systematic
approach to receiver design. There are four generic stages.

346

Chapter 15. Make It So

Sampled
received
signal
Downconversion

Matched
filter

Carrier
recovery

Interpolator
downsampler

Timing
recovery

Adaptive layer

Equalizer

Training
segment
locator

Equalizer
adaptation

Decision
device

Recovered
source
Decoder

Frame
synchronization

Figure 15.3 DSP portion of software-dened receiver.

1. Choose the order in which the basic operations of the receiver occur.
2. Select components and methods that can perform the basic operations in an
ideal setting.
3. Select adaptive elements that allow the receiver to continue functioning when
there are impairments.
4. Verify that the performance requirements are met.
While it may seem as though each stage requires that choices made in the preceding stages be xed, in reality, diculties encountered at one stage in the design
process may require a return to (and dierent choices to be made in) earlier
stages. As will soon become clear, the M6 problem specication has basically
(pre)resolved the design issues of the rst two stages.

15.2.1

Stage One: Ordering the Pieces

The rst stage is to select the basic components and the order in which they
occur. The design layout rst established in Figure 2.13 (and reappearing in the
DSP portion of the receiver in Figure 15.3) suggests one feasible structure. As the
signal enters the receiver it is downconverted (with carrier recovery), matched
ltered, interpolated (with timing recovery), equalized (adaptively), quantized,
and decoded (with frame synchronization). This classical ordering, while popular,
is not the only (nor necessarily the best) way to recover the message from the
noisy, ISI-distorted, FDMPAMIF received signal. However, it oers a useful
foundation for assessing the relative benets and costs of alternative receiver
congurations. Also, we know for sure that the M6 receiver can be built this
way. Other congurations may work, but we have not tested them.1
How was this ordering of components chosen? The authors have consulted
with, worked for, talked about (and argued with) engineers working on a number of receiver systems, including HDTV (high-denition television), DSL, and
1

If this sounds like a challenge, rest assured it is. Research continues worldwide, making
compilation of a complete handbook of receiver designs and algorithms a Sisyphean task.
The creation of new algorithms with minor variations that exploit a particular applicationspecic circumstance is a popular pastime of communication engineers. Perhaps you too will
come up with a unique approach!

Make It So

347

AlohaNet. The ordering of components in Figures 2.13 and 15.3 represents an

amalgamation of ideas from these (and other) systems. Sometimes it is easy to
argue why a particular order is good, sometimes it is a matter of preference or
personal experience, and sometimes the choice is based on factors outside the
engineers control.2
For example, the carrier-recovery algorithms of Chapter 10 are not greatly
aected by noise or intersymbol interference (as was shown in Exercises 10.36
and 10.40). Thus carrier recovery can be done before equalization, and this is the
path we have followed. But it need not be done in this order.3 Another example
is the placement of the timing-recovery element. The algorithms of Chapter 12
operate at baseband, and hence the timing recovery in Figure 15.3 is placed after
the demodulation. But there are passband timing-recovery algorithms that could
have been used to reverse the order of these two operations.

15.2.2

Stage Two: Selecting Components

Choices for the second design stage are relatively set as well. Since the sampling
is done at a sub-Nyquist rate fs (relative to the IF frequency fI ), the spectrum of the analog received signal is replicated every fs . The integer n for which
f = |fI nfs | is smallest denes the nominal frequency f from which further
downconversion is needed. Recall that such downconversion by sampling was discussed in Section 6.2. Using dierent specications, the M6 sampling frequency
fs may be above the Nyquist frequency associated with the IF frequency fI .4
The most common method of downconversion is to use mixing followed by
an FIR lowpass lter. This will be followed by an FIR matched lter, an
interpolatordecimator for downsampling, and a symbol-spaced FIR equalizer
that adapts its coecients in response to the training data contained in the transmission. The output of the equalizer is quantized to the nearest 4-PAM symbol
value, translated back into binary, decoded (using the (5, 2) block decoder), and
nally turned back into readable text.
Given adequate knowledge of the operating environment (the SNR in the
received signal, the carrier frequency and phase, the clock period and symbol
timing, and the marker location), the designer-selected parameters within these
components can be set to recover the message. This was, in fact, the strategy
followed in the idealized receiver of Chapter 9. Said another way, the choices
in stages one and two are presumed to admit an acceptable answer if properly tuned. Component selections at this point (including specication of the
2
3
4

For instance, the company might have a patent on a particular method of timing recovery
and using any other method might require royalty payments.
For instance, in the QAM radio of A Digital Quadrature Amplitude Modulation Radio, available in the website, the blocks appear in a dierent order.
Indeed, changing parameters such as this allows an instructor to create new transmission
standards for each class!

348

Chapter 15. Make It So

xed lowpass lter in the downconverter and the xed matched lter preceding
the interpolator/downsampler) can be conrmed by simulations of the ISI-free
ideal/full-knowledge setting. Thus, the upper half of Figure 15.3 is specied by
stage-two activities.

15.2.3

Stage Three: Anticipating Impairments

In the third design stage, the choices are less constrained. Elements of the third
stage are shown in the lower half of the receiver (the adaptive layer of Figure 15.3) and include the selection of algorithms for carrier, timing, frame synchronization, and equalizer adaptation. There are several issues to consider.
One of the primary stage-three activities is algorithm selectionwhich performance function to use in each block. For example, should the M6 receiver
use a phase-locked loop, a Costas loop, or a decision-directed method for carrier
recovery? Is a dual loop needed to provide adequate carrier tracking, or will a
single loop suce? What performance function should be used for the equalizer?
Which algorithm is best for the timing recovery? Is simple correlation suitable
to locate the training and marker segment?
Once the specic methods have been chosen, it is necessary to select specic
variables and parameters within the algorithms. This is a traditional aspect
of engineering design that is increasingly dominated by computer-aided design,
simulation, and visualization tools. For example, error surfaces and eye diagrams
can be used to compare the performance of the various algorithms in particular
settings. They can be used to help determine which technique is more eective
for the application at hand.
As software-aided design packages proliferate, the need to understand the computational mechanics underlying a particular design becomes less of a barrier.
For instance, Software Receiver Design has relied exclusively on the lterdesign algorithms built into Matlab. But the specication of the lter (its shape,
cuto frequencies, computational complexity, and lter length) cannot be left to
Matlab. The more esoteric the algorithm, the less transparent is the process
of selecting design parameters. Thus, Software Receiver Design has devoted
considerable space to the design and operation of adaptive elements.
But, even assuming that the trade-os associated with each of the individual
components are clear, how can everything be integrated together to succeed at
a multifaceted design objective such as the M6 receiver?

15.2.4

Sources of Error and Trade-os

Even when a receiver is fully operational, it might not decode every symbol
precisely. There is always a chance of error. Perhaps part of the error is due to a
frequency mismatch, part of the error is due to noise in the channel, part of the
error is due to a nonoptimal timing oset, etc. This section (and the next) suggest
a general strategy for allocating part of the error to each component. Then, as

Make It So

349

long as the sum of all the partial errors does not exceed the maximum allowable
error, there is a good chance that the complete receiver will work according to
its specications.
The approach is to choose a method of measuring the amount of error, for
instance, the average of the squared recovery error. Each individual component
can be assigned a threshold, and its parameters can be adjusted so that it does
not contribute more than its share to the total error. Assuming that the accumulation of the errors from various sources is additive, the complete receiver
will have no larger error than the concatenation of all its parts. This additivity
assumption is eectively an assumption that the individual pieces of the system
do not interact with each other. If they do (or when they do), then the threshold
allotments may need to be adjusted.
There are many factors that contribute to the recovery error, including the
following.

r Residual interference from adjacent FDM bands (caused by imperfect bandpass

r
r
r
r
r
r
r
r
r

ltering before downconversion and imperfect lowpass ltering after downconversion).

AGC jitter (caused by the deviation in the instantaneous signal from its
desired average and scaled by the stepsize in the AGC element).
Quantization noise in the sampler (caused by coarseness in the magnitudes of
the quantizer).
Round-o noise in lters (caused by wordlength limitations in lter parameters and lter algebra).
Residual interference from the doubly upconverted spectrum (caused by
imperfect lowpass ltering after downconversion).
Carrier phase jitter (occurs physically as a system impairment and is caused
by the stepsize in the carrier-recovery element).
Timing jitter (occurs physically as a system impairment and is caused by the
stepsize in the timing-recovery element).
Residual mean-squared error left by the equalizer (even an innitely long linear equalizer cannot remove all recovery error in the presence of simultaneous
channel noise and ISI).
Equalizer parameter jitter (caused by the nonvanishing stepsize in the adaptive equalizer).
Noise enhancement by the equalizer (caused by ISI that requires large equalizer gains, such as a deep channel null at frequencies that also include noise).

Because Matlab implements all calculations in oating-point arithmetic, the

quantization and round-o noise in the simulations is imperceptible. The project
setup presumes that the AGC has no jitter. A well-designed and suciently long
lowpass lter in the downconverter can eectively remove the interference from
outside the user band of interest. The in-band interference from sloppy adjacent
FDM signals should be considered part of the in-band channel noise. This leaves

350

Chapter 15. Make It So

carrier phase, timing jitter, imperfections in the equalizer, tap jitter, and noise
gain. All of these are potentially present in the M6 software-dened digital radio.
In all of the cases in which error is due to the jiggling of the parameters in
adaptive elements (in the estimation of the sampling instants, the phase errors,
the equalizer taps), the errors are proportional to the stepsize used in the algorithm. Thus, the (asymptotic) recovery error can be made arbitrarily small by
reducing the appropriate stepsize. The problem is that, if the stepsize is too
small, the element takes longer to converge. If the time to convergence of the element is too long (for instance, longer than the complete message), then the error
is increased. Accordingly, there is some optimal stepsize that is large enough to
allow rapid convergence yet small enough to allow acceptable error. An analogous trade-o arises with the choice of the length of the equalizer. Increasing its
length reduces the size of the residual error. But as the length grows, so does the
amount of tap jitter.
Such trade-os are common in any engineering design task. The next section
suggests a method of quantifying the trade-os to help make concrete decisions.

15.2.5

Tuning and Testing

The testing and verication stage of receiver design is not a simple matter
because there are so many things that can go wrong. (There is so much stu
that can happen!) Of course, it is always possible to simply build a prototype and then test to see whether the specications are met. Such a haphazard
approach may result in a working receiver, but then again, it might not. Surely
there is a better way! This section suggests a commonsense approach that is not
uncommon among practicing engineers. It represents a practical compromise
between excessive analysis (such as one might nd in some advanced communication texts) and excessive trial and error (such as try something and cross
your ngers).
The idea is to construct a simulator that can create a variety of test signals that
fall within the M6 specication. The parameters within the simulator can then
be changed one at a time, and their eect on various candidate receivers noted.
By systematically varying the test signals, the worst components of the receiver
can be identied and then replaced. As the tests proceed, the receiver gradually
improves. As long as the complete set of test signals accurately represents the
range of situations that will be encountered in operation, the testing will lead to
a successful design.
Given the particular stage-one and -two design choices for the M6 receiver,
the previous section outlined the factors that may degrade the performance of
the receiver. The following steps suggest some detailed tests that may facilitate
the design process.

Make It So

351

r Step 1: Tuning the Carrier Recovery As shown in Chapter 10, any of the

carrier-recovery algorithms will be capable of locating a xed phase oset in

a receiver in which everything else is operating optimally. Even when there is
noise or ISI, the best settings for the frequency and phase of the demodulation
sinusoid are those that match the frequency and phase of the carrier of the
IF signal. For the M6 receiver, there are two issues that must be considered.
First, the M6 specication allows the frequency to be (somewhat) dierent
from its nominal value. Is a dual-loop structure needed? Or can a single loop
adequately track the expected variations? Second, the transmitter phase may
be jittering.
The user-choosable features of the carrier-recovery algorithms are the LPF
and the algorithm stepsize, both of which inuence the speed at which the
estimates can change. Since the carrier-recovery scheme needs to track a timevarying phase, the stepsize cannot be chosen too small. Since a large stepsize
increases the error due to phase jitter, it cannot be chosen too large. Thus,
an acceptable stepsize will represent a compromise.
To conduct a test to determine the stepsize (and LPF) requires creating test
signals that have a variety of o-nominal frequency osets and phase jitters.
A simple way to model phase jitter is to add a lowpass-ltered version of
zero-mean white noise to a nominal value. The quality of a particular set
of parameters can then be measured by averaging (over all the test signals)
the mean squared recovery error. Choosing the LPF and stepsize parameters
to make this error as small as possible gives the best values. This average
error provides a measure of the portion of the total error that is due to the
carrier-recovery component in the receiver.
Step 2: Tuning the Timing Recovery As shown in Chapter 12, there are several
algorithms that can be used to nd the best timing instants in the ideal setting.
When the channel impairment consists purely of additive noise, the optimal
sampling times remain unchanged, though the estimates will likely be more
noisy. As shown by Example 12.3, and in Figure 12.12, however, when the
channel contains ISI, the answer returned by the algorithms diers from what
might be naively expected.
There are two parts to the experiments at this step. The rst is to locate the
best timing-recovery parameter for each test signal. (This value will be needed
in the next step to assess the performance of the equalizer.) The second part
is to nd the mean squared recovery error due to jitter of the timing-recovery
algorithm.
The rst part is easy. For each test signal, run the chosen timing-recovery
algorithm until it converges. The convergent value gives the timing oset
(and indirectly species the ISI) to which the equalizer will need to respond.
(If it jiggles excessively, then decrease the stepsize.)
Assessing the mean squared recovery error due to timing jitter can be done
much like the measurement of jitter for the carrier recovery: measure the
average error that occurs over each test signal when the algorithm is initialized

352

Chapter 15. Make It So

at its optimum answer; then average over all the test signals. The answer may
be aected by the various parameters of the algorithm: the that determines
the approximation to the derivative, the l parameter that species the time
support of the interpolation, and the stepsize (these variable names are from
the rst part of the timing-recovery algorithm clockrecDD.m on page 258.)
In operation, there may also be slight inaccuracies in the specication of the
clock period. When the clock period at the transmitter and that at the receiver
dier, the stepsize must be large enough that the timing estimates can follow
the changing period. (Recall the discussion surrounding Example 12.4.) Thus,
again, there is a tension between a large stepsize needed to track rapid changes
and a small stepsize to minimize the eect of the jitter on the mean squared
recovery error. In a more complex environment, in which clock phases might be
varying, it might be necessary to follow a procedure more like that considered
in step 1.
Step 3: Tuning the Equalizer After choosing the equalizer method (as specied
by the performance function), there are parameters that must be chosen and
decisions that must be made in order to implement the linear equalizer. These
are
the order of the equalizer (number of taps),
the initial values of the equalizer,
the training signal delay (if using the training signal), and
the stepsize.
As in the previous steps, it is a good idea to create a collection of test signals
using a simulation of the transmitter. To test the performance of the equalizer, the test signals should contain a variety of ISI channels and/or additive
interferences.
As suggested in Chapter 13, in a high-SNR scenario the T -spaced equalizer
tries to implement an approximation of the inverse of the ISI channel. If the
channel is mild, with all its roots well away from the unit circle, then its
inverse may be fairly short. But if the channel has zeros that are near the unit
circle, then its FIR inverse may need to be quite long. While much can be
said about this, a conservative guideline is that the equalizer should be from
two to ve times longer than the maximum anticipated channel delay spread.
One subtlety that arises in making this decision and in consequent testing is
that any channel ISI that is added into a simulation may appear dierently
at the receiver because of the sampling. This eect was discussed at length in
Section 13.1, where it was shown how the eective digital model of the channel
includes the timing oset. Thus (as mentioned in the previous step) assessing the actual channel to which the equalizer will adapt requires knowing
the timing oset that will be found by the timing recovery. Fortunately, in
the M6 receiver structure of Figure 15.3, the timing-recovery algorithm operates independently of the equalizer, and so the optimal value can be assessed
beforehand.

Make It So

353

For most of the adaptive equalizers in Chapter 13, the center-spike initialization is used. This was justied in Section 13.4 (see page 289) as a useful
method of initialization. Only if there is some concrete a priori knowledge of
the channel characteristics would other initializations be used.
The problem of nding an appropriate delay was discussed in Section 13.2.3,
where the least-squares solution was recomputed for each possible delay. The
delay with the smallest error was the best. In a real receiver, it will not be
possible to do an extensive search, and so it is necessary to pick some delay.
The M6 receiver uses correlation to locate the marker sequence and this can be
used to locate the time index corresponding to the rst training symbol. This
location plus half the length of the equalizer should correspond closely to the
desired delay. Of course, this value may change depending on the particular
ISI (and channel lengths) used in a given test signal. Choose a value that,
over the complete set of test signals, provides a reasonable answer.
The remaining designer-selected variable is stepsize. As with all adaptive
methods, there is a trade-o inherent in stepsize selection: making it too large
can result in excessive jitter or algorithm instability, while making it too small
can lead to an unacceptably long convergence time. A common technique is
to select the largest stepsize consistent with achievement of the components
assigned asymptotic performance threshold.
Step 4: Frame Synchronization Any error in locating the rst symbol of each
four-symbol block can completely garble the reconstructed text. The frame
synchronizer operates on the output of the quantizer, which should contain few
errors once the equalizer, timing recovery, and phase recovery have converged.
The success of frame synchronization relies on the peakiness of the correlation
of the marker/training sequence. The chosen marker/training sequence A0Oh
well whatever Nevermind should be long enough that there are few false
spikes when correlating to nd the start of the message within each block.
To test software written to locate the marker, feed it a sample symbol string
assembled according to the specications described in the previous section as
if the downconverter, clock timing, equalizer, and quantizer had recovered the
transmitted symbol sequence perfectly.

Finally, after tuning each component separately, it is necessary to conrm

that, when all the pieces of the system are operating simultaneously, there are
no excessive negative interactions. With luck, little further tuning will prove
necessary to complete a successful design. The next section has more specics
about the M6 receiver design.
Exercise 15.1. From the parameters in Table 15.1, do the following.
a. Determine the oversampling (or upsampling) factor.
b. Determine what frequency the sampled received signal will be centered
around. Recall that the receiver front end samples at a rate lower than the IF
frequency, and your answer should be something between 0 and fs /2.

354

Chapter 15. Make It So

Exercise 15.2. The B 3IG transmitter can be used to implement the M6 transmission specications, as provided in m6params.m. Plot the received signal and
its spectrum. Use these plots to explain why your answers to Exercise 15.1 are
correct.

15.3

No Soap Radio: The M6 Receiver Design Challenge

The analog front end of the receiver in Figure 15.2 takes the signal from an
antenna, amplies it, and crudely bandpass lters it to (partially) suppress frequencies outside the desired users frequency band. An analog downconverter
modulates the received signal (approximately) down to the nominal IF fI at
2 MHz. The output of the analog downconverter is set by an automatic gain
controller to t the range of the sampler. The output of the AGC is sampled at
intervals of Ts = 850 kHz to give r[k], which provides a Nyquist bandwidth
of 425 kHz that is ample for a 102 kHz baseband user bandwidth. The sampled
received signal r[k] from Figure 15.2 is the input to the DSP portion of the
receiver in Figure 15.3.
The following comments on the components of the digital receiver in Figure 15.3 help characterize the design task.

r The downconversion to baseband uses the sampler frequency fs , the known

r
r

r
r

intermediate frequency fI , and the current phase estimates to determine the

mixer frequency needed to demodulate the signal. The M6 receiver may use
any of the phase-tracking algorithms of Chapter 10. A second loop may also
help with frequency oset.
The lowpass ltering in the demodulator should have a bandwidth of roughly
102 kHz, which will cover the selected source spectrum but reject components
outside the frequency band of the desired user.
The interpolator/downsampler implements the reduction in sample rate to
T -spaced values. This block must also implement the timing synchronization,
so that the time between samples after timing recovery is representative of
the true spacing of the samples at the transmitter. You are free to implement
this in any of the ways discussed in Chapter 12.
Since there could be a signicant amount of intersymbol interference due
to channel dynamics, an equalizer is essential. Any one will do. A trained
equalizer requires nding the start of the marker/training segment while a
blind equalizer may converge more slowly.
The decision device is a quantizer dened to reproduce the known alphabet
of the s[i] by a memoryless nearest-element decision.
At the nal step, the decoding from blockcode52.m in conjunction with
bin2text.m can be used to reconstruct the original text. This also requires
a frame synchronization that nds and removes the start block consisting of

Make It So

355

marker plus training, which is most likely implemented using a correlation

technique.
The software-dened radio should have the following user-selectable variables
that can readily be set at the start of processing of the received block of data:

r
r
r
r

rollo
initial
initial
initial

factor for the square-root raised cosine pulse shape,

phase oset,
timing oset, and
equalizer parameterization.

The following are some suggestions.

r Use the B 3IG transmitter with m6params.m to implement the M6 transmis-

sion specication along with various impairments. This will enable you to test
your receiver (as described in the preceding section) over a wider range of
conditions than just the test signals available on the website.
Try to break your receiver. See how much noise can be present in the received
signal before accurate (e.g., less than 1% symbol errors) demodulation seems
impossible. Find the fastest change in the carrier phase that your receiver can
track, even with a bad initial guess.
In order to facilitate more eective debugging while building the project,
implementation of a debug mode in the receiver is recommended. The information of interest will be plots of the time histories of pertinent signals as well
as timing information (e.g., a graph of matched lter average output power
versus receiver symbol timing oset). One convenient way to add this feature
to your Matlab receiver would be to include a debug ag as an argument
that produces these plots when the ag is activated.
When debugging adaptive components, use a test with initialization at the
right answer and zero stepsize to determine whether the problem is in the
adaptation or in the xed component structure. An initialization very near
the desired answer with a small stepsize will reveal that the adaptive portion
is working properly if the adaptive parameter trajectory remains in the close
vicinity of the desired answer. A rapid divergence may indicate that the update
has the wrong sign or that the stepsize is way too large. An aimless wandering
that drifts away from the vicinity of the desired answer represents a more
subtle problem that requires reconsideration of the algorithm code and/or its
suitability for the circumstance at hand.

Several test les that contain a mystery signal with a quote from a
well-known book are available on the website. They are labeled easy.mat,
medium.mat, and hard.mat.5 These have been created with a variety of dierent
5

One student remarked that these should have been called hard.mat, harder.mat, and
completelyridiculous.mat. Nonetheless, a well-crafted M6 receiver can recover the hidden
messages.

356

Chapter 15. Make It So

rollo factors, carrier frequencies, phase noises, ISI, interferers, and symbol timing osets. We encourage the adventurous reader to try to receive these secret
signals. Solve the mystery. Break it down.

For Further Reading

An overview of a practical application of software-dened radio emphasizing the
redenability of the DSP portion of the receiver can be found in

r B. Bing and N. Jayant, A Cellphone for All Standards, IEEE Spectrum, pp.
3439, May 2002.
The eld of software radio erupted with a special issue of the IEEE Communications Magazine in May 1995. This was called a landmark special issue in
an editorial in the more recent

r J. Mitola, III, V. Bose, B. M. Leiner, T. Turletti, and D. Tennenhouse, Eds.,

IEEE Journal on Selected Areas in Communications (Special Issue on Software Radios), vol. 17, April 1999.
For more information on the technological context and the relevance of software
implementations of communications systems, see

r E. Buracchini, The Software Radio Concept, IEEE Communications Magazine, vol. 38, pp. 138143, September 2000
and papers from the (occasional) special section in the IEEE Communications
Magazine on topics in software and DSP in radio. For much more, see

r J. H. Reed, Software Radio: A Modern Approach to Radio Engineering,

Prentice-Hall, 2002,
which overlaps in content (if not style) with the rst half of Software Receiver
Design.
Two recommended monographs that devote more attention than most to the
methodology of the same slice of digital receiver design as we consider here are

r J. A. C. Bingham, The Theory and Practice of Modem Design, Wiley Interscience, 1988 (especially Chapter 5);

r H. Meyr, M. Moeneclaey, and S. A. Fechtel, Digital Communication Receivers:

Synchronization, Channel Estimation, and Signal Processing, Wiley Interscience, 1998 (especially Section 4.1).

16 A Digital Quadrature Amplitude

Modulation Radio

The preceding chapters of Software Receiver Design focus on a real-valued

PAM transmission protocol. Though this is one of the simplest digital transmission systems, a working radio for PAM needs to include a variety of techniques that address the nonideal behaviors encountered in reality. Fixes such
as the PLL, the Costas loop, timing recovery, and equalization are necessary
parts of the system. This chapter describes a complex-valued generalization of
PAM called quadrature amplitude modulation (QAM), which forms the basis
of a better radio. While the change from real-valued PAM to complex-valued
QAM signals is straightforward, the bulk of this chapter eshes out the details
of operation of the resulting system. The various parts of the digital radio need
to be rethought (carrier recovery, timing recovery, and equalization). The theme
of this chapter is that such rethinking can follow the same path as the original
designs. The basic techniques for the PAM radio of earlier chapters generalize
smoothly to more sophisticated designs.
This chapter begins with the observation (already noted in Section 5.3) that
PAM modulations waste bandwidth. QAM, a more clever modulation strategy,
is able to halve the bandwidth requirements, which represents a signicant step
towards practicality. This is accomplished by sending, and then unravelling, two
real PAM signals simultaneously on orthogonal carriers (sine and cosine) that
occupy the same frequency band. A straightforward way to describe QAM is to
model the two signals as the real and imaginary portions of a single complexvalued signal.

16.1

The Song Remains the Same

Sections 16.2 and 16.3 show explicitly how QAM can be viewed as PAM that uses
complex-valued modulation and demodulation. In an ideal world, this would be
the end of the story. But reality intrudes in several ways. Just as a PAM radio
needs to align the phase at the transmitter with the phase at the receiver, so

358

Chapter 16. A Digital Quadrature Amplitude Modulation Radio

will a QAM radio. While the basic ideas of carrier recovery remain the same,
the details of the QAM protocol require changes in the operation of the adaptive
elements used to estimate the parameters. Similarly, the frequency of the carrier
at the transmitter and the frequency of the demodulating signal at the receiver
must be aligned. For example, Section 16.4.1 extends the Costas-loop idea to
4-QAM while Section 16.4.3 extends the PLL for 4-QAM.
Just as a PAM radio needs to align the sampling times and rates at the receiver
with the symbol times and rates, so will a QAM radio. While the basic ideas
of timing recovery remain the same, the details of the QAM protocol require
changes in the detailed operation of the adaptive elements. Section 16.6 extends
the timing-recovery methods of PAM in Chapter 12 to QAM.
Just as a PAM radio needs to build an equalizer to combat intersymbol interference, so will a QAM radio. While the basic ideas of equalization remain the
same as in Chapter 13, equalization of the two real signal streams of QAM is most
easily described in terms of a complex-valued adaptive equalizer as in Section
16.8.
An interesting feature of QAM is that the carrier recovery can be carried
out before the equalizer (as in Section 16.4) or after the equalizer. Such postequalization carrier recovery can be accomplished as in Section 16.7 by derotating
the complex-valued constellation at the output of the equalizer. This is just one
example of the remarkable variety of ways in which a QAM receiver can be
structured. Several dierent possible QAM constellations are discussed in Section
16.5 and several possible ways of ordering the receiver elements are drawn from
the technical literature in Section 16.9.
As is done throughout Software Receiver Design, exercises are structured
as mini-design exercises for individual pieces of the system. These pieces come
together at the end, with a description of (and Matlab code for) the Q3 AM
prototype receiver, which parallels the M6 receiver of Chapter 15, but for QAM.
A series of experiments with the Q3 AM receiver is intended to encourage exploration of a variety of alternative architectures for the radio. Exercises ask for
assessments of the practicality of the various extensions, and allow for experimentation, discovery, and perhaps even a small dose of frustration, as the importance of real-world impairments stresses the radio designs. The objective of this
chapter is to show the design process in action and to give the reader a feel for
how to generalize the original PAM system for new applications.

16.2

Quadrature Amplitude Modulation (QAM)

Sections 5.1 and 5.2 demonstrate the use of amplitude modulation with and
without a (large) carrier. An unfortunate feature of these modulation techniques
(and of the resulting real-valued AM systems) is that the range of frequencies
used by the modulated signal is twice as wide as the range of frequencies in the
baseband. For example, Figures 4.10 and 5.3 (on pages 70 and 82) show baseband

359

A Digital Quadrature Amplitude Modulation Radio

signals that are nonzero for frequencies between B and B while the spectrum
of the modulated signals is nonzero in the interval [fc B, fc + B] and in
the interval [fc B, fc + B]. Since bandwidth is a scarce resource, it would be
preferable to use a scheme such as quadrature modulation (QM, introduced in
Section 5.3), which provides a way to remove this redundancy. Since the spectrum
of the resulting signal is asymmetric in QM, such a system can be modeled using
complex-valued signals, though these can always be rewritten and realized as (a
pair of) real-valued signals.
As described in Equation (5.6) on page 91, QM sends two message signals
m1 (t) and m2 (t) simultaneously in the same 2B passband bandwidth using two
orthogonal carriers (cosine and sine) at the same frequency fc ,
v(t) = m1 (t) cos(2fc t + ) m2 (t) sin(2fc t + ).

(16.1)

is the xed oset of the carrier phase. Dene the complex-valued message
m(t) = m1 (t) + jm2 (t)
and the complex sinusoid ej(2fc t+) . Observe that the real part of the product
m(t)ej(2fc t+) = m1 (t) cos(2fc t + ) m2 (t) sin(2fc t + )
+ j[m2 (t) cos(2fc t + ) + m1 (t) sin(2fc t + )]
is precisely v(t) from (16.1), that is,
v(t) = Re{m(t)ej(2fc t+) },

(16.2)

where Re{a + jb} = a takes the real part of its argument.

The two message signals may be built (as in previous chapters) by pulseshaping a digital symbol sequence using a time-limited pulse shape p(t). Thus
the messages are represented by the signal

si [k]p(t kT ),
(16.3)
mi (t) =
k

where T is the interval between adjacent symbols and si [k], drawn from a nite
alphabet, is the ith message symbol at time k. The corresponding transmitted
signal is

v(t) =
p(t kT )[s1 [k] cos(2fc t + ) s2 [k] sin(2fc t + )]
k

= Re

#
p(t kT )s[k]e

j(2fc t+)

(16.4)

where
s[k] = s1 [k] + js2 [k]
is a complex-valued symbol obtained by combining the two real-valued messages.
The two data streams could arise from separate messages. For instance, one
user might send a binary 1 sequence s1 [k] while another user transmits a binary

360

Chapter 16. A Digital Quadrature Amplitude Modulation Radio

11

s2
1

1
10

01

s1

1
1
(a)

00

01

s2

1
11

00

s1

1
1

10

Figure 16.1 Two choices for associating

symbol pairs (s1 , s2 ) with bit pairs. In
(a), (1, 1) is associated with 00, (1, 1)
with 01, (1, 1) with 11, and (1, 1)
with 10. In (b), (1, 1) is associated
with 10, (1, 1) with 00, (1, 1) with 01,
and (1, 1) with 11.

(b)

1 sequence s2 [k]. Taken together, these can be thought of as the 4-QAM constellation 1 j. There are many ways in which two binary signals can be combined;
two are shown in Figure 16.1, which plots (s1 , s2 ) pairs in a two-dimensional plane
with s1 values on one axis and s2 values on the other.
Alternatively, the two messages could be a joint encoding of a single message. For example, a 1, 3 4-PAM coding could be recast as 4-QAM using an
assignment like
3 1 j
1 1 + j
+1 +1 j
+3 +1 + j
In either case, s1 [k] and s2 [k] are assumed to have zero average and to be uncorrelated so that their average product is zero.
The process of generating QAM signals such as (16.4) is demonstrated in
qamcompare.m, which draws on the PAM pulse-shaping routine pulrecsig.m
from page 194. qamcompare.m uses a simple Hamming pulse shape p(t) and random binary values for the messages s1 [k] and s2 [k]. The modulation is carried out
in two ways, with real sines and cosines (corresponding to the rst part of (16.4))
and with complex modulation (corresponding to the second part of (16.4)). The
two outputs can be compared by calculating max(abs(vcomp-vreal)), which
veries that the two are identical.
Listing 16.1. qamcompare.m compare real and complex QAM implementations
N=1000; M=20; Ts = . 0 0 0 1 ;
% # symbols , o v e r s a m p l i n g f a c t o r
time=Ts * (N*M 1 ) ; t =0: Ts : time ;
% s a m p l i n g i n t e r v a l and time
s 1=pam (N, 2 , 1 ) ; s 2=pam (N, 2 , 1 ) ;
% l e n g t h N r e a l 2 l e v e l s i g n a l s
ps=hamming (M) ;
% p u l s e shape o f width M
f c =1000; th = 1.0; j=sqrt ( 1 ) ;
% c a r r i e r f r e q and phase
s1up=zeros ( 1 ,N*M) ; s2up=zeros ( 1 ,N*M) ;
s1up ( 1 :M: end)= s 1 ;
% o v e r s a m p l e by M
s2up ( 1 :M: end)= s 2 ;
% o v e r s a m p l e by M
sp1=f i l t e r ( ps , 1 , s1up ) ;
% c o n v o l v e p u l s e shape with s 1
sp2=f i l t e r ( ps , 1 , s2up ) ;
% c o n v o l v e p u l s e shape with s 2
% make r e a l and complexv a l u e d v e r s i o n s and compare
v r e a l=sp1 . * cos (2 * pi * f c * t+th ) sp2 . * s i n (2 * pi * f c * t+th ) ;
vcomp = r e a l ( ( sp1+j * sp2 ) . * exp ( j * (2 * pi * f c * t+th ) ) ) ;
max( abs ( vcompv r e a l ) )
% v e r i f y t h a t they r e th e same

A Digital Quadrature Amplitude Modulation Radio

0000

0001
3

0100

0101
1

s2

0011

0010

0111

0110

1100

1
1101

1111

1110

1000

3
1001

1011

1010

361

Figure 16.2 A Gray-coded 16-QAM assigns bit

values to symbol values so that the binary
representation of each symbol diers from its
neighbors by exactly one bit. The corresponding
16-QAM constellation contains the values
s = s1 + js2 for s1 and s2 1, 3.

s1

It is also possible to use larger QAM constellations. For example, Figure 16.2
shows a Gray-coded 16-QAM constellation that can be thought of as the set of
all possible s = s1 + js2 , where s1 and s2 take values in 1, 3. Various other
possibilities are discussed in Section 16.5.
The two message signals in QAM may also be built using a pulsed-phase
modulation (PPM) method rather than the pulsed-amplitude method described
above. Though they are implemented somewhat dierently, both can be written
in the same form as (16.4). A general PPM sequence can be written

p(t kT ) cos(2fc t + (t)),
(16.5)
v(t) = g
k

where g is a xed scaling gain and (t) is a time-varying phase signal dened by
the data to be transmitted. For example, a 4-PPM system might have
(t) = [k],

kT t < (k + 1)T,

(16.6)

where [k] is chosen from among the four possibilities /4, 3/4, 5/4, and
7/4. The four phase choices are associated with the four pairs 00, 10, 11, and
01 to convert from the bits of the message bits to the transmitted signal. This
particular phase modulation is sometimes called quadrature phase-shift keying
(QPSK) in the literature.
Using the cosine angle-sum formula (A.13),
cos(2fc t + (t)) = cos(2fc t) cos((t)) sin(2fc t) sin((t)).

With g = 2,
g cos([k]) = 1 and g sin([k]) = 1,
for all values of [k]. Hence
cos(2fc t + (t)) = cos(2fc t) sin(2fc t))

362

Chapter 16. A Digital Quadrature Amplitude Modulation Radio

and the resulting QPSK signal can be written


p(t kT )[s1 [k] cos(2fc t) s2 [k] sin(2fc t)],
v(t) =
k

which is the same as (16.4) with si = 1 and with zero phase = 0. Thus, QPSK
and QAM are just dierent realizations of the same underlying process.
Exercise 16.1. Use qamcompare.m as a basis to implement a 4-PPM modulation.
Identify the values of si that correspond to the values of [k]. Demonstrate that
the two modulated signals are the same.
Exercise 16.2. Use qamcompare.m as a basis to implement a 16-QAM modulation as shown in Figure 16.2. Compare this system with 4-QAM in terms
of
a. amount of spectrum used, and
b. required power in the transmitted signal.
What other trade-os are made when moving from a 4-QAM to a 16-QAM
modulation scheme?
Exercise 16.3. Consider a complex-valued modulation system dened by
v(t) = Im{m(t)ej(2fc t+) },
where Im{a + jb} = b takes the imaginary part of its argument.
a. Find a modulation formula analogous to the left-hand side of (16.4) to show
that v(t) be implemented using only real-valued signals.
b. Demonstrate the operation of this imaginary modulation scheme in a simulation modeled on qamcompare.m.
Exercise 16.4. Consider the pulsed-phase modulation scheme in (16.5)
(16.6) with 16 elements, where the angle can take values = 2n/16 for
n = 0, 1, . . . , 15. How does this scheme dier from, and how is it similar to, the
16-QAM of Figure 16.2?
Exercise 16.5. In PAM, the data are modulated by a single sinusoid. In QAM,
two sets of data are modulated by a pair of sinusoids of the same frequency but
90 out of phase. Is it possible to build a communications protocol in which four
sets of data are modulated by four sinusoids 45 out of phase? If so, describe the
modulation procedure. If not, explain why not.

363

A Digital Quadrature Amplitude Modulation Radio

e j(2f0t+)

e j(2fct+)
m(t)
Re{ }

v(t)

x(t)

s(t)
LPF

Transmitter

Receiver

Figure 16.3 Quadrature-modulation transmitter and receiver (complex-valued

version). When f0 = fc and = , the output s(t) is equal to the message m(t). This
is mathematically the same as the real-valued system shown in Figure 16.5.

16.3

Demodulating QAM
Figure 5.10 on page 91 shows how a quadrature-modulated signal like (16.1) can
be demodulated using two mixers, one with a sin() and one with a cos(). The
underlying trigonometry (under the assumptions that the carrier frequency fc
is known exactly and that the phase term is zero) is written out in complete
detail in Section 5.3, and is also shown (in the context of nding the envelope
of a complex signal) in Appendix C. This section focuses on the more practical
case where there may be phase and frequency osets. The math is simplied by
using the complex-valued form where the received signal v(t) of (16.2) is mixed
with the complex sinusoid ej(2f0 t+) as in Figure 16.3.
Algebraically, this can be written
x(t) = ej(2f0 t+) v(t) = ej(2f0 t+) Re{m(t)ej(2fc t+) }
= ej(2f0 t+) (m1 (t) cos(2fc t + ) m2 (t) sin(2fc t + ))
=

ej()
(m1 (t)ej2(f0 +fc )t + m1 (t)ej2(f0 fc )t )
2
ej()
(m2 (t)ej2(f0 +fc )t + m2 (t)ej2(f0 fc )t ).
+j
2

Assuming fc f0 and that the cuto frequency of the LPF is less than 2fc (and
greater than fc ), the two high-frequency terms are removed by the LPF so that
s(t) = LPF{x(t)} =
=

ej()
(m1 (t)ej2(f0 fc )t + jm2 (t)ej2(f0 fc )t )
2

1 j2(f0 fc )t+j()
e
m(t),
2

(16.7)

where m(t) = m1 (t) + jm2 (t). When the demodulating frequency f0 is identical
to the modulating frequency fc and when the modulating phase is identical to
the demodulating phase , the output of the receiver s(t) is an attenuated copy
of the message m(t). The spectra of the various signals in the demodulation
process are shown in Figure 16.4. The transmitted signal v(t) is real-valued and
hence has a symmetric spectrum. This is shifted by the mixer so that one of the
(asymmetric) pieces of |X(f )| is near the origin and the other lies near 2fc.
The lowpass lter removes the high-frequency portion, leaving only the copy of

364

Chapter 16. A Digital Quadrature Amplitude Modulation Radio

|V( f )|

|X( f )|

fc

2fc

fc

fc

|S( f )|

fc

fc

fc

Figure 16.4 Spectra of the various signals in the complex-valued demodulator of

Figure 16.3. Observe the asymmetry of the spectra |X(f )| and |S(f )|.
cos(2fct+)

cos(2f0t+)

m1(t)

x1(t)

LPF

s1(t)

Transmitter
m2(t)

sin(2fct+)

v(t)

Receiver
x2(t)

LPF

s2(t)

sin(2f0t+)

Figure 16.5 Quadrature modulation transmitter and receiver (real-valued version).

This is mathematically identical to the system shown in Figure 16.3.

m(t) near the origin. Though the calculations use complex-valued signals, the
same procedure can be implemented purely with real signals as shown in Figure
16.5.
When f0 = fc but = , s(t) of (16.7) simplies to ej() m(t)/2. This can be
interpreted geometrically by observing that multiplication by a complex exponential ej is a rotation (in the complex plane) through an angle . In terms of
the pulsed data sequences s1 and s2 shown in Figures 16.1 and 16.2, the constellations become rotated about the origin through an angle given by . This
is the dierence between the actual and the estimated phase. In order to recover
the transmitted data properly (to deduce the correct sequence of symbols), it is
necessary to derotate s(t). This can be accomplished by estimating the unknown
values using techniques that are analogous to the phase-locked loop, the Costas
loop, and decision-directed methods, as will be discussed in Section 16.4. The
details will change in order to account for the dierences in the problem setting,
but the general techniques for carrier recovery will be familiar from the PAM
setting in Chapter 10.
When f0 = fc and = , the output of the demodulator in (16.7) is s(t) =
1
j2(f0 fc )t
. This is a time-varying rotation of the constellation: the points
2 m(t)e
s1 and s2 spin as a function of time at a rate proportional to the frequency
dierence between the carrier frequency fc and the estimate f0 . Again, it is
necessary to estimate this dierence using carrier-recovery techniques.

A Digital Quadrature Amplitude Modulation Radio

365

This complex-valued modulation and demodulation is explored in qamdemod.m.

A message m is chosen randomly from the constellation 1 j and then pulseshaped with a Hamming blip to form mp. Modulation is carried out as in (16.2)
to give the real-valued v. Demodulation is accomplished by a complex-valued
mixer as in Figure 16.3 to give x, which is then lowpass ltered. Observe that
the spectra of the signals (as shown using plotspec.m, for instance) agree with
the stylized drawings in Figure 16.4. After the lowpass ltering, the demodulated
s is equal to the original mp, though it is shifted in time (by half the length of
the lter) and attenuated (by a factor of 2).
Listing 16.2. qamdemod.m modulate and demodulate a complex-valued QAM signal
N=10000; M=20; Ts = . 0 0 0 1 ;
time=Ts * (N*M 1 ) ; t =0: Ts : time ;
m=pam (N, 2 , 1 ) + j *pam (N, 2 , 1 ) ;
ps=hamming (M) ;
f c =1000; th = 1.0;
mup=zeros ( 1 ,N*M) ; mup ( 1 :M: end)=m;
mp=f i l t e r ( ps , 1 , mup ) ;
v=r e a l (mp. * exp ( j * (2 * pi * f c * t+th ) ) ) ;
f 0 =1000; ph = 1.0;
x=v . * exp( j * (2 * pi * f 0 * t+ph ) ) ;
l =50; f = [ 0 , 0 . 2 , 0 . 2 5 , 1 ] ;
a =[1 1 0 0 ] ; b=f i r p m ( l , f , a ) ;
s=f i l t e r ( b , 1 , x ) ;

%
%
%
%
%
%
%
%
%
%
%
%
%

# symbols , o v er s a m p l e M
sampling times
s i g na l s of length N
p u l s e shape o f width M
c a r r i e r f r e q and phase
o v e r s a m p l e by M
c o n v o l v e p u l s e shape & data
complex c a r r i e r
f r e q and phase o f demod
demodulate v
s p e c i f y f i l t e r p a r a m e te r s
design f i l t e r
s=LPF{ x }

Exercise 16.6. qamdemod.m shows that when fc = f0 and = the reconstruction of the message is exact. The symbols in m can be recovered if the real and
imaginary parts of s are sampled at the correct time instants.
a. Find the correct timing oset tau so that s[k+tau] = m[k] for all k.
b. Examine the consequences of a phase oset. Let = 0 (th=0) and examine
the output when (ph) is 10 , 20 , 45 , 90 , 180 , and 270 . Explain what
you see in terms of (16.7).
c. Examine the consequences of a frequency oset. Let fc=1000, f0=1000.1, and
th=ph=0. What is the correct timing oset tau for this case? Is it possible to
reconstruct m from s? Explain what you see in terms of (16.7).
Exercise 16.7. Implement the demodulation shown in Figure 16.5 using realvalued sines and cosines to carry out the same task as in qamdemod.m. (Hint:
begin with vreal of qamcompare.m and ensure that the message can be recovered
without use of complex-valued signals.)
Exercise 16.8. Exercise 16.2 generates a 16-QAM signal. What needs to be
changed in qamdemod.m to successfully demodulate this 16-QAM signal?
Exercise 16.9. Design a complex-valued demodulation for the imaginary
method of Exercise 16.3. Show that the demodulation can also be carried out
using purely real signals.

366

Chapter 16. A Digital Quadrature Amplitude Modulation Radio

ej(2f0t+)
v(t)

phase
splitter

x(t)

s(t)

Figure 16.6 An alternative demodulation scheme

for QAM uses a phase splitter followed by a
mixing with a complex exponential.

Exercise 16.10. Multiply the complex number a + jb by ej . Show that the

answer can be interpreted as a rotation through an angle . Verify that the
matrix


cos() sin()
R() =
sin() cos()
 
a
carries out the same rotation when multiplied by the vector
.
b
Exercise 16.11. Let s1 (t), s2 (t), m1 (t), and m2 (t) be dened as the inputs and
outputs of Figure 16.5. Show that




1
s1 (t)
m1 (t)
= R()
s2 (t)
m2 (t)
2
for = 2(fc f0 )t + , where the rotation matrix R() is dened in Exercise 16.10.
Exercise 16.12. Is it possible to build a demodulator for the 16-PPM protocol
of Exercise 16.4? Draw a block diagram or explain why it cannot be done.
Exercise 16.13. Is it possible to build a demodulator for the four-cosine modulation scheme of Exercise 16.5? Draw a block diagram or explain why it cannot
be done.
Exercise 16.14. An alternative method of demodulating a QAM signal uses a
phase splitter, which is a lter with Fourier transform

1 f 0,
(f ) =
(16.8)
0 f < 0.
The phase splitter is followed by exponential multiplication as in Figure 16.6.
a. Implement a phase-splitter demodulation directly in the frequency domain
using (16.8).
b. Sketch the spectra of the signals v(t), x(t), and s(t) in Figure 16.6. The answer
should look something like the spectra in Figure 16.4.
c. Show that when fc = f0 and = the reconstruction of the message is exact.
d. Examine (by simulation) the consequences of a phase oset.
e. Examine (by simulation) the consequences of a frequency oset.

A Digital Quadrature Amplitude Modulation Radio

16.4

367

Carrier Recovery for QAM

It should be no surprise that the ability to reconstruct the message symbols from
a demodulated QAM signal depends on identifying the phase and frequency of
the carrier, because these are also required in the simpler PAM case of Chapter
10. Equation (16.7) illustrates explicitly what is required in order to recover
m(t) from s(t). To see how the recovery might begin, observe that the receiver
can adjust the phase and demodulating frequency f0 to any desired values. If
necessary, the receiver could vary the parameters over time. This section discusses
two methods of building an adaptive element to iteratively estimate a timevarying , including the Costas loop for 4-QAM (which parallels the discussion
in Section 10.4 of the Costas loop for PAM) and a phase-locked loop for 4-QAM
(which parallels the discussion in Section 10.3 of a PLL for PAM). The variations
in phase can then be parlayed into improved frequency estimates following the
same kinds of strategies as in Section 10.6.

16.4.1

Costas Loop for 4-QAM

If it were possible to choose a time-varying (t) in (16.7) so that
(t) = 2(fc f0 )t +

(16.9)

then s(t) would be directly proportional to m(t) because ej2(f0 fc )t+j((t)) =

1. The simplest place to begin is to assume that the frequencies are correct (i.e.,
that fc = f0 ) and that is xed but unknown.
As in previous chapters, the approach is to design an adaptive element by
specifying a performance (or objective) function. Equation (16.9) suggests that
the goal in carrier recovery of QAM should involve an adaptive parameter that
is a function of the (unknown) parameter . Once this objective has been chosen,
an iterative algorithm can be implemented to achieve convergence.
Sampling perfectly downconverted 4-QAM signals s1 (t) and s2 (t) at the proper
times should produce one of the four pairs (1, 1), (1, 1), (1, 1), and (1, 1)
from the symbol constellation diagram in Figure 16.1. A rotation through 90
(for any integer ) will also produce samples with the same values, though they
may be a mixture of the symbol values from the two original messages. (Recall
that Exercise 16.6(b) asked for an explicit formula relating the various rotations
to the sampled (s1 , s2 ) pairs.)
For the PAM system of Chapter 10, the phase is designed to converge to an
oset of an integer multiple of 180 . The corresponding objective function JC for
the Costas loop from (10.15) is proportional to cos2 ( ), which has maxima
at = + for integers . This observation can be exploited by seeking an
objective function for QAM that causes the carrier-recovery phase to converge
to an integer multiple of 90 . It will later be necessary to resolve the ambiguity
among the four choices. This is postponed until Section 16.4.2.

368

Chapter 16. A Digital Quadrature Amplitude Modulation Radio

Consider the objective function

JCQ = cos2 (2 2),

(16.10)

which can be rewritten using the cosine-squared formula (A.4) as

1
(1 + cos(4 4)).
2
JCQ achieves its maximum whenever cos(4 4) = 1, which occurs whenever
4( ) = 0, 2, 4, . . .. This is precisely
JCQ =

= + (/2)
for integers , as desired. The objective is to adjust the receiver mixer phase (t)
so as to maximize JC . The result will be the correct phase or a 90 , 180 , or
270 rotation of .
The adaptive update to maximize JCQ is

JC 
.
(16.11)
[k + 1] = [k] +
=[k]
In order to implement this, it is necessary to generate

JC
|=[k] ,

JC
= 2 sin(4 4).

which is
(16.12)

The next step is nd a way to generate a signal proportional to sin(4 4) for

use in the algorithm (16.11) without needing to know .
The analogous derivative for the PAM Costas loop of Section 10.4 is the product of LPF{v(t) cos(2f0 t + )} and LPF{v(t) sin(2f0 t + )}, where v(t) is the
received signal. Since these are demodulations by two signals of the same frequency but 180 out of phase, a good guess for the QAM case is to lowpass lter
v(t) times four signals, each 90 out of phase, and to take the product of these
four terms. The next several equations show that this is a good guess since this
product is directly proportional to the required derivative sin(4 4).
The received signal for QAM is v(t) of (16.1). Dene the four signals
x1 (t) = LPF {v(t) cos (2f0 t + )} ,
+

,
x2 (t) = LPF v(t) cos 2f0 t + +
,
4 ,
+


,
x3 (t) = LPF v(t) cos 2f0 t + +
2 


3
.
x4 (t) = LPF v(t) cos 2f0 t + +
4

A Digital Quadrature Amplitude Modulation Radio

369

Using (A.8) and (A.9), and assuming f0 = fc ,


1
m1 (t)[cos( ) + cos(4fc t + + )]
x1 (t) = LPF
2

m2 (t)[sin( ) + sin(4fc t + + )]

1
[m1 (t) cos( ) m2 (t) sin( )].
2

Similarly,
1
[m1 (t) cos( /4) m2 (t) sin( /4)] ,
2
1
x3 (t) = [m1 (t) sin( ) + m2 (t) cos( )] ,
2
1
x4 (t) = [m1 (t) cos( 3/4) m2 (t) sin( 3/4)] .
2
Now form the product of the four xi (t)s and simplify to obtain
x2 (t) =

x1 (t)x2 (t)x3 (t)x4 (t) =

1
[4m1 (t)m2 (t)(m22 (t) m21 (t)) cos(4 4) (16.13)
128
(m41 (t) 6m21 (t)m22 (t) + m42 (t)) sin(4 4)].

For a pulse shape p(t) such as a rectangle or the Hamming blip that is timelimited to within one T -second symbol interval, p(t) = 0 for t < 0 or t > T .
Squaring both sides of the QAM pulse train (16.3) gives

m2i (t) =
s2i [k]p2 (t kT )
k

p2 (t kT ) (t)

for 4-QAM, where si [k] = 1 and (t) is some periodic function with period T .
Thus,
m21 (t) m22 (t) = 0,
m41 (t) = m42 (t) = m21 (t)m22 (t) = 2 (t).
Accordingly, the four-term product simplies to
x1 (t)x2 (t)x3 (t)x4 (t) = 42 (t) sin(4 4),
which is directly proportional to the desired derivative (16.12).
In this special case (where the pulse shape is time-limited to a single symbol
interval), of (16.11) becomes

[k + 1] = [k] + x1 (t)x2 (t)x3 (t)x4 (t) t=kTs ,=[k] ,
(16.14)
where Ts denes the time between successive updates (which is typically less
than the symbol period T ).

370

Chapter 16. A Digital Quadrature Amplitude Modulation Radio

LPF

x1[k]

cos(2f0kTs +[k])

cos(2f0kTs +[k] + /4)

LPF

x2[k]

v(kTs)

LPF

x3[k]

[k]

Figure 16.7 This quadriphase

Costas-loop adaptive element
can be used to identify an
unknown phase oset in a
4-QAM transmission. The
gure is a block diagram of
Equation (16.14).

cos(2f0kTs +[k] + /2)

cos(2f0kTs +[k] + 3/4)

LPF

x4[k]

This adaptive element is shown schematically in Figure 16.7, where the

received signal v(t) is sampled to give v[k] and then mixed with four out-ofphase cosines. Each path is lowpass ltered (with a cuto frequency above 2fc )
and then multiplied to give the update term. The integrator is the standard
structure of an adaptive element, indicating that the new value is updated at
each iteration with a step size of .
The QAM Costas loop of (16.14) and Figure 16.7 is implemented in
qamcostasloop.m. The received signal is from the routine qamcompare.m on page
360, where the variable vreal is created by modulating a pair of binary message
signals. The four lters are implemented using the time-domain method of
Section 7.2.1 and the entire procedure mimics the Costas loop for PAM, which
is implemented in costasloop.m on page 208.
Listing 16.3. qamcostasloop.m simulate Costas loop for QAM
r=v r e a l ;
% v r e a l from qamcompare .m
f l =100; f =[0 . 2 . 3 1 ] ; a =[1 1 0 0 ] ;
% f i l te r specification
h=f i r p m ( f l , f , a ) ;
% LPF d e s i g n
mu= . 0 0 3 ;
% algorithm s t e p s i z e
f 0 =1000; q= f l +1;
% f r e q at r e c e i v e r
th=zeros ( 1 , length ( t ) ) ; th (1)=randn ;
% i n i t i a l i z e estimate
z1=zeros ( 1 , q ) ; z2=zeros ( 1 , q ) ;
% i n i t i a l i z e LPFs
z4=zeros ( 1 , q ) ; z3=zeros ( 1 , q ) ;
% z s contain past inputs
f o r k =1: length ( t )1
s =2* r ( k ) ;
z1 =[ z1 ( 2 : q ) , s * cos (2 * pi * f 0 * t ( k)+ th ( k ) ) ] ;
z2 =[ z2 ( 2 : q ) , s * cos (2 * pi * f 0 * t ( k)+ pi/4+ th ( k ) ) ] ;
z3 =[ z3 ( 2 : q ) , s * cos (2 * pi * f 0 * t ( k)+ pi/2+ th ( k ) ) ] ;
z4 =[ z4 ( 2 : q ) , s * cos (2 * pi * f 0 * t ( k)+3 * pi/4+ th ( k ) ) ] ;

371

A Digital Quadrature Amplitude Modulation Radio

Phase offset

Figure 16.8 Depending on where it is

initialized, the estimates made by the
QAM Costas-loop element converge to
/2. For this plot, the
unknown was 1.0, and there
were 50 dierent initializations.

4
0

0.2

0.4

0.6

Time

l p f 1=f l i p l r ( h ) * z1 ;
l p f 2=f l i p l r ( h ) * z2 ;
l p f 3=f l i p l r ( h ) * z3 ;
l p f 4=f l i p l r ( h ) * z4 ;
th ( k+1)= th ( k)+mu* l p f 1 * l p f 2 * l p f 3 * l p f 4 ;
end

%
%
%
%
%

output o f f i l t e r
output o f f i l t e r
output o f f i l t e r
output o f f i l t e r
a l g o r i t h m update

1
2
3
4

Typical output of qamcostasloop.m appears in Figure 16.8, which shows the

evolution of the phase estimates for 50 dierent starting values of , th(1). Some
of these converge to the correct phase oset = 1.0, and others converge to
nearby values 1 + /2 for integers . As expected, these stationary points
occur at all the 90 osets of the desired value of .
When the frequency is not exactly known, the phase estimates of the Costas
algorithm try to follow. For example, in Figure 16.9, the frequency of the carrier
is f0 = 1000, while the assumed frequency at the receiver was fc = 1000.1. Fifty
dierent starting points were used, and in all cases the estimates converge to
lines with the same slope. The slope of these lines is directly proportional to
the dierence in frequency between fc and f0 , and this slope can be used as in
Section 10.6 to estimate the frequency dierence.
Exercise 16.15. Use the preceding code to play with the Costas loop for
QAM.
a.
b.
c.
d.

How does the stepsize mu aect the convergence rate?

What happens if mu is too large (say mu=1)?
Does the convergence speed depend on the value of the phase oset?
When there is a small frequency oset, what is the relationship between the
slope of the phase estimate and the frequency dierence?

Exercise 16.16. How does the lter h inuence the performance of the Costas
loop?

Chapter 16. A Digital Quadrature Amplitude Modulation Radio

2
Phase offset

372

Figure 16.9 When the frequency of the

carrier is unknown at the receiver, the
phase estimates converge to lines.
The slope of these lines is
proportional to the frequency oset,
and can be used to correct the
estimate of the carrier frequency.

6
0

0.2

0.4

0.6

Time

a. Try fl = 1000, 30, 10, 3. In each case, use freqz to verify that the lter has
been properly designed.
b. Remove the LPFs completely from costasloop.m. How does this aect the
convergent values? The tracking performance?
Exercise 16.17. What changes are needed to qamcostasloop.m to enable it to
function with 16-QAM? (Recall Exercises 16.2 and 16.8.)
Exercise 16.18. Exercise 10.21 shows an alternative implementation of the PAM
Costas loop that uses cheaper xed-phase oscillators rather than adjustablephase oscillators. Refer to Figure 10.12 on page 211. Design an analogous QAM
implementation and write a simulation (or modify qamcostasloop.m) to demonstrate its operation.
Exercise 16.19. Figure 16.10 shows a high-level block diagram of a dual-loop
system for tracking a frequency oset in a QAM system.
a. Draw a detailed block diagram labeling all signals and blocks. For comparison,
the PAM equivalent (using PLLs) is shown in Figure 10.18.
b. Implement the dual-loop frequency tracker and verify that small frequency
osets can be tracked. For comparison, the PAM equivalent is implemented
in dualplls.m on page 219.
c. How large can the frequency oset be before the tracking fails?
The previous discussion focused on the special case where the pulse shape is
limited to a single symbol interval. In fact, the same algorithm functions for more

A Digital Quadrature Amplitude Modulation Radio

v[k]

Carrier
Recovery

Carrier
Recovery

373

1[k]

2[k]

[k]

Figure 16.10 Dual-carrier-recovery-loop

Conguration.

general pulse shapes. The key is to nd conditions under which the update term
(16.13) remains proportional to the derivative of the objective function (16.12).
The generic eect of any small-stepsize adaptive element like (16.14) is to
average the update term over time. Observe that the four-term product (16.13)
has two parts: one is proportional to cos(4 4) and the other is proportional
to sin(4 4). Two things must happen for the algorithm to work: the average
of the cos term must be zero and the average of the sin term must always have
the same sign. The average of the rst term is
avg{4m1 (t)m32 (t) 4m2 (t)m31 (t)}
while the average of the second is
avg{m41 (t) 6m21 (t)m22 (t) + m42 (t)}.
In QAM, the messages mi (t) are built as the product of the data si [k] and a
pulse shape as in (16.3). Even assuming that the average of the data values
is zero (which implies that the average of the messages is also zero), nding
these by hand for various pulse shapes is not easy. But they are straightforward
to nd numerically. The following program calculates the desired quantities by
averaging over N symbols. The default pulse shape is the Hamming blip, but
others can be tested by redening the variable unp=srrc(10,0.3,M,0) for a
square-root raised cosine pulse or unp=ones(1,M) for a rectangular pulse shape.
In all cases, the cosine term cterm is near zero while the sine term sterm is a
negative constant. Since the term appears in (16.13) with a negative sign, the
sine term is proportional to the derivative sin(4 4). Thus it is possible to
approximate all the terms that enter into the algorithm (16.14).
Listing 16.4. qamconstants.m calculate constants for avg{x1 x2 x3 x4}
N=2000; M=50;
% # o f symbols , o v er s a m p l e M
s 1=pam (N, 2 , 1 ) ; s 2=pam (N, 2 , 1 ) ;
% r e a l 2 l e v e l s i g n a l s
mup1=zeros ( 1 ,N*M) ; mup2=zeros ( 1 ,N*M) ;
mup1 ( 1 :M: end)= s 1 ;
% z e r o pad Ts p a c e d s e q u e n c e
mup2 ( 1 :M: end)= s 2 ;
% o v e r s a m p l e by M
unp=hamming (M) ;
% u n n o r m a l i z e d p u l s e shape
p=sqrt (M) * unp / sqrt (sum( unp . 2 ) ) ;
% n o r m a l i z e d p u l s e shape
m1=f i l t e r ( p , 1 , mup1 ) ;
% c o n v o l v e p u l s e shape & data
m2=f i l t e r ( p , 1 , mup2 ) ;
% c o n v o l v e p u l s e shape & data
sterm=sum( ( m1. 4 ) 6 * (m1 . 2 ) . * ( m2. 2 ) + (m2 . 4 ) ) / (N*M)
cterm=sum(4 * m1 . * m2.3 4 * m1 . 3 . * m2) / (N*M)

374

Chapter 16. A Digital Quadrature Amplitude Modulation Radio

Exercise 16.20. Find a pulse shape p(t) and a set of symbol values s1 and s2
that will cause the QAM Costas loop to fail. Hint: nd values so that the average
of the sinusoidal term is near zero. Redo the simulation qamcostasloop.m with
these values and verify that the system does not achieve convergence.
Exercise 16.21. Suppose that si [k] are binary 1 and that avg{s1 [k]} =
avg{s2 [k]} = 0. The message sequences mi (t) are generated as in (16.3). Show
that avg{m1 (t)} = avg{m2 (t)} = 0 for any pulse shape p(t).
Exercise 16.22. Suppose that avg{m1 (t)} = avg{m2 (t)} = 0. Which of the
following are true?
a.
b.
c.
d.
e.
f.

avg{m1 (t)} + avg{m2 (t)} = avg{m1 (t) + m2 (t)}

avg{4m1 (t)} + 2 avg{m2 (t)} = avg{4m1 (t) + 2m2 (t)}
avg{m1 (t)m2 (t)} = avg{m1 (t)} avg{m2 (t)}
avg{m21 (t)} = avg{m1 (t)}2
avg{m21 (t)m22 (t)} = avg{m21 (t)} avg{m22 (t)}
Show that, even if (c) holds, (e) need not.

Other variations of this quadriphase Costas loop can be found in Sections

6.3.1 and 6.3.2 of Bingham (1988) and in Section 4.2.3 of Anderson (1999).

16.4.2

Resolution of the Phase Ambiguity

The carrier-recovery element for QAM does not have a unique convergent
point because the objective function JCQ of (16.10) has maxima whenever
= + /2. This is shown clearly in the multiple convergence points of Figure
16.8. The four possibilities are also evident from the rotational symmetry of the
constellation diagrams (such as Figures 16.1 and 16.2). There are several ways
in which the phase ambiguity can be resolved, including
1. the output of the downsampler can be correlated with a known training signal,
2. the message source can be dierentially encoded so that the information is
carried in the successive changes in the symbol values (rather than in the
symbol values themselves)
One way to implement the rst method is to correlate the real part of the
training signal t with both the real s1 and the imaginary s2 portions of the
output of the downsampler. Whichever correlation has the larger magnitude is
selected and its polarity examined. To write this concisely, let t, si  represent

A Digital Quadrature Amplitude Modulation Radio

375

the correlation of t with si . The four possibilities are as follows:

if |t, s1 | > |t, s2 |
t, s1  > 0 no change,
t, s1  < 0 add 180;
if |t, s1 | < |t, s2 |
t, s2  > 0 add 90 ,
t, s2  < 0 add 90 .
Using dierential encoding of the data bypasses the need for resolving the
ambiguous angle. To see how this might work, consider the 2-PAM alphabet 1
and re-encode the data so that a +1 is transmitted when the current symbol is
the same as the previous symbol and a 1 is transmitted whenever it changes.
Thus the sequence
1, 1, 1, 1, 1, 1, 1, 1, 1, . . .
is dierentially encoded (starting from the second element) as
1, 1, 1, 1, 1, 1, 1, 1, . . .
The process can be reversed as long as the value of the rst bit (in this case a 1)
is known. This strategy can be extended to complex-valued 4-QAM by encoding
the dierences s[k] s[k 1]. This is described in Figure 16-4 of Barry, Lee, and
Messerschmitt (2004). An advantage of dierential encoding is that no training
sequence is required. A disadvantage is that a single error in the transmission
can cause two symbol errors in the recovered sequence.
Exercise 16.23. Implement the correlation method to remove the phase ambiguity.
a. Generate a complex 4-QAM signal with a phase rotation by use of
qamcompare.m. Specify a xed training sequence.
b. Use the QAM Costas loop qamcostasloop.m to estimate .
c. Implement the correlation method in order to disambiguate the received signal. Verify that the disambiguation works with = 60 , 85 , and 190 .
Exercise 16.24. Design a dierential encoding scheme for 4-QAM using successive dierences in the data. Verify that the scheme is insensitive to a 90
rotation.

16.4.3

Phase-Locked Loop for 4-QAM

Section 10.1 shows how the carrier signal in a PAM transmission is concealed
within the received signal. Using a squaring operation followed by a bandpass

376

Chapter 16. A Digital Quadrature Amplitude Modulation Radio

lter (as in Figure 10.3 on page 197) passes a strong sinusoid at twice the frequency and twice the phase of the carrier. This double-frequency surrogate can
then be used as an input to a phase-locked loop for carrier recovery as in Section
10.3.
In QAM, the carrier is better hidden because it is generated as in (16.1) by
combining two dierent signals modulated by both sin() and cos() or, equivalently, by the real and imaginary parts of a complex modulating wave as in
(16.2). What kind of preprocessing can bring the carrier into the foreground?
An analogy with the processing in the Costas loop suggests using the fourth
power of the received signal. This will quadruple (rather than double) the frequency of the underlying carrier, and so the fourth power needs to be followed
by a narrow-bandpass lter centered at (about) four times the carrier frequency.
This quadruple-frequency surrogate can then be used as an input to a standard
phase-locked loop for carrier recovery.
The rst step is to verify that using the square really is futile. With v(t) dened
in (16.1), it is straightforward (though tedious) to show that
v 2 (t) =

1 2
[m (t) + m22 (t) + (m21 (t) m22 (t)) cos(4fc t + 2)
2 1
2m1 (t)m2 (t) sin(4fc t + 2)].

A bandpass lter with center frequency at 2fc removes the constant DC term,
and this cannot be used (as in PAM) to extract the carrier since avg{m21 (t)
m22 (t)} = avg{m1 (t)m2 (t)} = 0. Nothing is left after passing v 2 (t) through a
narrow-bandpass lter.
The fourth power of v(t) has three kinds of terms: DC terms, terms near
2fc , and terms near 4fc . All but the latter are removed by a BPF with center
frequency near 4fc . The remaining terms are
BPF{v 4 (t)} =

1
[4m1 (t)m2 (t)(m22 (t) m21 (t)) sin(8fc t + 4)
8
(m41 (t) 6m21 (t)m22 (t) + m42 (t)) cos(8fc t + 4)].

What is amazing about this equation is that the two terms 4m1 m2 (m22 m21 )
and m41 6m21 m22 + m42 are identical to the two terms in (16.13). Accordingly,
the former averages to zero while the latter has a nonzero average. Thus
BPF{v 4 (t)} cos(8fc t + 4)
and this signal can be used to identify (four times) the frequency and (four times)
the phase of the carrier.
In typical application, the carrier frequency is known only approximately, that
is, the BPF must use the center frequency f0 rather than fc itself. In order to
make sure that the lter passes the needed sinusoid, it is wise to make sure that
the bandwidth is not too narrrow. On the other hand, if the bandwidth of the
BPF is too wide, unwanted extraneous noises will also be passed. This tension

A Digital Quadrature Amplitude Modulation Radio

v(kTs)

X4

BPF
centered
at ~ 4fc

LPF

377

[k]

sin(8f0kTs +4[k])

Figure 16.11 The sampled QAM input v(kTs ) is raised to the fourth power and passed

through a BPF. This extracts a sinusoid with frequency 4fc and phase 4. A
standard PLL extracts 2 for use in carrier recovery.

suggests a design trade-o between the uncertainty in fc and the desire for a
narrow BPF.
Exercise 16.25. Generate a complex 4-QAM signal v(t) with phase shift as in
qamcompare.m. Imitate the code in pllpreprocess.m on page 196, but raise the
signal to the fourth power and choose the center frequency of the BPF to be 4fc.
Verify that the output is (approximately) a cos wave of the desired frequency
4fc and phase 4. This may be more obvious in the frequency domain.
Exercise 16.26. What happens on raising v(t) to the eighth power? Can this
be used to recover (eight times) the carrier frequency at (eight times) the phase?
Explain why or why not.
The output BPF{v4 (t)} of the preprocessor is presumed to be a sinusoid that
can be applied to a phase-tracking element such as the digital phase-locked loop
(PLL) from Section 10.3. The preprocessor plus PLL are shown in Figure 16.11,
which is the same as the PLL derived for PAM in Figure 10.7, except that it
must operate at 4fc (instead of 2fc ) and extract a phase equal to 4 (instead
of 2). Observe that this requires that the digital signal be sampled at a rate
sucient to operate at 4fc in order to avoid aliasing. In practice, such a rapid
sampling rate can be a tall order.
Though this section has focused on a fourth-power method of carrier recovery
for 4-QAM, the methods are also applicable to larger constellations such as 16QAM. The literature oers some encouraging comments. For instance, Anderson
(1999) on page 149 states:
When the . . . modulation is, for instance 16-QAM instead of QPSK, the PLL reference signal is not a steady cos(40 t + 40 ) and the phase dierence signal has other
components besides sin(40 40 ). Nonetheless, the synchronizer still works passably
well.

Bingham (1988) on page 173 says:

This simple method [PLL tracking of quadrupled frequency] can be used for constellations with 16 points . . . but it has been generally assumed that the pattern jitter
would be intolerable. However, it can be shown that, at least for 16-QAM, the outer
points dominate, and the fourth-power signal has a component at 4fc that is usable
if a very narrow band PLL is used to rene it. Whether the independence from data

378

Chapter 16. A Digital Quadrature Amplitude Modulation Radio

decisions and equalizer convergence that this forward-acting method oers outweighs
the problems of such a narrow-band PLL remains to be seen.

Exercise 16.27. Implement the 4-QAM PLL system shown in Figure 16.11.
Begin with the preprocessor from Exercise 16.25 and combine this with a modied version of pllconverge.m from page 204.
Exercise 16.28. What changes need to be made to the 4-QAM preprocessor of
Exercise 16.25 and the corresponding PLL of Exercise 16.27 in order to operate
with a 16-QAM constellation?

16.5

Designing QAM Constellations

This chapter has focused on 4-QAM, and occasionally considered other QAM
constellations such as the 16-QAM conguration in Figure 16.2. This section
shows a few common variations and suggests some of the trade-os that occur
with dierent spacing of the points in the constellation.
As with PAM, the clearest indicator of the resilience of a constellation with
respect to noise is the distance between points. The transmission is less likely to
have bits (and hence symbols) changed when the points are far apart and more
likely to encounter errors when the points are close together. This means that,
when greater power is used in the transmission, more noise can be tolerated.
Accordingly, it is common to normalize the power and to consider the symbol
error rate (SER) as a function of the signal-to-noise ratio (SNR). Eectively,
this means that the points in a constellation such as 16-QAM of Figure 16.2 are
closer together than in a simpler constellation such as 4-QAM.
It is straightforward to calculate the SER as a function of the SNR via simulation. The routine qamser.m species a M-QAM constellation and then generates
N symbols. Noise with variance N0 is added and then a comparison is made. If
the noise causes the symbol to change then an error has occurred. This can be
plotted for a number of dierent noise variances. The most complicated lines
in the routine are those comparing the real and imaginary portions of the data
s and the noisy version r to determine whether the noise has caused a symbol
error.
Listing 16.5. qamser.m symbol error rate for QAM in additive noise
N0 = 1 0 . ( 0 : . 2 : 3 ) ;
% noise variances
N=10000;
% # symbols to s i m u l a t e
M=4;
% # symbols i n c o n s t e l l a t i o n
s=pam (N, sqrt (M) , 1 )+ j * pam (N, sqrt (M) , 1 ) ; % QAM symbols
s=s / sqrt ( 2 ) ;
% n o r m a l i z e power
c o n s t=u n i q u e ( r e a l ( s ) ) ;
% n o r m a l i z e d symbol v a l u e s
a l l s e r s = zeros ( s i z e (N0 ) ) ;
f o r i =1: length (N0)
% l o o p o v e r SNR
n=sqrt (N0( i ) / 2 ) * ( randn ( 1 ,N)+ j * randn ( 1 ,N ) ) ;
r=s+n ;
% r e c e i v e d s i g n a l+n o i s e

A Digital Quadrature Amplitude Modulation Radio

Symbol Error Rate

10

10

10

10

Figure 16.12 SER versus SNR for

16-QAM
10

379

256-QAM
64-QAM

4-QAM
0

10

15
20
SNR in dB

25

various M -QAM constellations.

For a given SNR, simpler
constellations provide a more
robust transmission. For a given
constellation, larger SNRs give
fewer errors. The black dots are
plotted by the simulation in
qamser.m. The lines are a plot of
the theoretical formula (16.15).

30

r e a l e r r=q u a n ta l p h ( r e a l ( r ) , c o n s t )==q u a n ta l p h ( r e a l ( s ) , c o n s t ) ;
i m a g e r r=q u a n t a l p h ( imag( r ) , c o n s t )==q u a n ta l p h ( imag( s ) , c o n s t ) ;
SER=1mean( r e a l e r r . * i m a g e r r ) ;
% d e t e r m i n e SER by c o u n t i n g
a l l s e r s ( i )=SER ;
% # symbols w/ RE+IM c o r r e c t
end

When the percentage of errors is plotted against the SNR, the result is a
waterfall plot such as Figure 16.12. Large noises (small SNR) imply large
error rates while small noises cause few errors. Thus each constellation shows a
characteristic drop-o as the SNR increases. On the other hand, for a given SNR,
the constellations with the most widely distributed points have fewer errors.
For Gaussian noises (as generated by randn), soft errors in recovering the message pairs are circularly distributed about each constellation point. Accordingly,
it is wise to try to keep the points as far apart as possible, under the constraint
of equal power. Thus any four-point constellation should be a square in order to
avoid one pair of symbol errors becoming more likely.
The simulation qamser.m can be run with any noise distribution and with
any conceivable constellation. For regular QAM constellations and independent
Gaussian noise, there is a formula that predicts the theoretical shape of the
waterfall curve. Let a2 be the variance of the Gaussian noise and s2 the variance
of the symbol sequence. Proakis and Salehi (2002), on page 422, show that the
probability of an incorrect decision for an M -QAM system is
'%2
&
$


3s2
1
erfc
,
(16.15)
1 1 1
2(M 1)a2
M
where the erfc is the complementary error function (see help erfc in Matlab).
This formula is superimposed on the simulations for 4-, 16-, 64-, and 256-QAM
in Figure 16.12 and obviously agrees closely with the simulated results. This plot
was generated by the routine sersnr.m which is available on the website.
Of course, there is no need to restrict a constellation to a regular grid of points,
though rectangular grids will tend to have better noise rejection. For example,

380

Chapter 16. A Digital Quadrature Amplitude Modulation Radio

Figure 16.13 The V.29 constellation has 16 symbols

and four dierent energy levels.

a system designer could choose to omit the four corner points in a square QAM
constellation. This would reduce the signal power range over which analog electronics of the system must retain linearity. Nonsquare QAM constellations can
also be easier to synchronize, though they will tend to have higher SER (for
the same SNR) than square constellations because there is a smaller minimum
distance between points in constellations of the same total power. For example, Figure 16.13 shows an irregular constellation called the V.29 constellation.
Anderson (1999) comments on page 97 that the V.29 modem standard
has worse error performance in AWGN than does . . . [square] 16-QAM, but it is easier
to synchronize.

Another consideration in constellation design is the assignment of bit values

to the points of the constellation. One common goal is to reduce the number of
bit changes between adjacent symbols. For 4-QAM, this might map 45 00,
135 01, 135 11, and 45 10. Mappings from constellation points to
symbols that minimize adjacent symbol errors are called Gray codes.
Exercise 16.29. Adapt qamser.m to draw the waterfall plot for the V.29 constellation of Figure 16.13.
Exercise 16.30. If dierent noises are used, the waterfall plot may look significantly dierent than Figure 16.12. Redraw the SER vs SNR curves for 4-, 16-,
64-, and 256-QAM when the noise is uniform.

16.6

Timing Recovery for QAM

When a QAM signal arrives at the receiver it is a complicated analog waveform
that must be demodulated and sampled in order to eventually recover the two
messages hidden within. The timing-recovery element species the exact times
at which to sample the demodulated signals. One approach would be to use
two timing parameters 1 and 2 operating independently on the two message
signals s1 and s2 using any of the methods described in Chapter 12. But the two

381

A Digital Quadrature Amplitude Modulation Radio

sampled signals are not unrelated to each other. Though the messages themselves
are typically considered to be uncorrelated, the two signals have been modulated
and demodulated simultaneously, and so they are tightly synchronized with each
other. A better approach is to adapt a single parameter that species when to
sample both signals.
The problem is approached by nding performance or objective functions that
have maxima (or minima) at the optimal point (i.e., at the correct sampling
instants). The performance functions are then used to dene adaptive elements
that iteratively estimate the correct sampling times. As usual in Software
Receiver Design, all other aspects of the system are presumed to operate
awlessly: the upconversions and downconversions are ideal, the pulse-matched
lters are well designed, there is no interference, and the channel is benign. Moreover, both baseband signals are assumed to occupy the same bandwidth and the
sampling is fast enough to avoid problems with aliasing.
The oversampled sequences are ltered by matched lters that correspond to
the pulse shape at the transmitter. The output of the matched lters is sent into
interpolators that extract the two baud-spaced sequences. When all goes well,
these can be decoded into the original messages s1 and s2 . For PAM signals as in
Chapter 12, the baud-timing rate can be found by optimizing the average of the
rst, second, or fourth power of the absolute values of the baud-spaced sequence
values. For some pulse shapes and powers, it is necessary to minimize, whereas
others require maximization.
The demodulated QAM signal contains two independent signals s1 and s2 with
the same timing oset, or, equivalently, it is a single complex-valued signal s. By
analogy with PAM, it is reasonable to consider optimizing the average of powers
of s, that is, to maximize or minimize
JQ ( ) = avg{(|s1 |p + |s2 |p )q }

(16.16)

for p = 1, 2, or 4 and q = 12 , 1, or 2. Of course, other values for p and q might

also be reasonable.
While the error surfaces for these objective functions are dicult to draw
exactly, they are easy to calculate numerically, as is familiar for PAM from
clockrecDDcost.m on page 259 (also see Figures 12.8 and 12.11). The routine
qamtimobj.m calculates the value of the objective function over a range of different s for a given pulse shape unp. In this simulation, an oset of = 12 is the
desired answer.
Listing 16.6. qamtimobj.m draw error surfaces for baseband signal objective functions
m=20; n =1000;
unp=hamming (m) ;
cp=conv ( unp , unp ) ;
ps=sqrt (m) * cp / sqrt (sum( cp . 2 ) ) ;
c o s t=zeros (m+ 1 , 1 ) ;
f o r i =1:m+1
s 1=pam ( n/m, 2 , 1 ) ; s 2=pam ( n/m, 2 , 1 ) ;
mup1=zeros ( 1 , n ) ; mup1 ( 1 :m: end)= s 1 ;

%
%
%
%

n data p o i n t s , m+1 t a u s
p u l s e shape
matched f i l t e r
normalize

% f o r each o f f s e t
% c r e a t e 4QAM s e q u e n c e
% z e r o pad and upsample

Chapter 16. A Digital Quadrature Amplitude Modulation Radio

Value of objective functions

382

(c)

(e)
(b)
(d)
(a)
0

0.2

0.4
0.6
Timing offset

0.8

Figure 16.14 Five objective functions

JQ ( ) of (16.16) for the timing-recovery
problem are plotted versus the timing
oset . For the computation, = 12 is
the desired answer. This gure assumes
the Hamming pulse shape. (a) p = 2,
q = 12 , (b) p = 2, q = 1, (c) p = 2, q = 2,
(d) p = 1, q = 1, (e) p = 4, q = 1.

mup2=zeros ( 1 , n ) ; mup2 ( 1 :m: end)= s 2 ;

m1=f i l t e r ( ps , 1 , mup1 ) ;
% c o n v o l v e p u l s e shape & data
m2=f i l t e r ( ps , 1 , mup2 ) ;
% c o n v o l v e p u l s e shape & data
sm1=m1 ( ( length ( ps )1)/2+ i :m: end ) ;
% sampled baseband data
sm2=m2 ( ( length ( ps )1)/2+ i :m: end ) ;
% with t i m i n g o f f s e t iT /m
c o s t ( i )=sum( sqrt ( sm1.2+sm2 . 2 ) ) / length ( sm1 ) ; % o b j . f u n c t i o n
end

The output of qamtimobj.m is shown in Figure 16.14 for several objective

functions using the Hamming pulse shape:
*
(a) p = 2, q = 12 : avg{ s21 + s22 }
(b) p = 2, q = 1: avg{s21 + s22 }
( c) p = 2, q = 2: avg{(s21 + s22 )2 }
(d) p = 1, q = 1: avg{|s1 | + |s2 |}
( e) p = 4, q = 1: avg{s41 + s42 }
Observe that the solid curve (c), called the fourth-power objective function
because it can also be written avg{||s||4 }, has the steepest slope about the desired
answer. This should exhibit the fastest convergence, though it will also react most
quickly to noises. This is an example of the classic trade-o in adaptive systems
between speed of convergence and robustness to disturbances. Similar curves can
be drawn for other pulse shapes such as the square-root raised cosine. As occurs
for the PAM case in Figure 12.11, some of these may require maximization, while
others require minimization.
Using the fourth-power objective function (e) (i.e., p = 4, q = 1 in (16.16)), an
adaptive element can be built by descending the gradient. This is

(m41 (kT + ) + m42 (kT + )) 

[k + 1] = [k]


= [k]

= [k]
(m31 (kT + ) + m32 (kT + ))

(m1 (kT + ) + m2 (kT + )) 
.

= [k]

A Digital Quadrature Amplitude Modulation Radio

383

As in Sections 12.3 and 12.4, the derivative can be approximated numerically

[k + 1] = [k] (m31 (kT + [k]) + m32 (kT + [k]))
[m1 (kT + + ) m1 (kT + )
+ m2 (kT + + ) m2 (kT + )],
where =
/ and is small and positive. All of the values for the mi (t) for
the osets [k], [k] , and [k] + can be interpolated from the oversampled
mi . Clearly, a similar derivation can be made for any of the objective functions
(a)(e).
Which criterion is best? Barry, Lee, and Messerschmitt (2004) on page 747
say that:
For some signals, particularly when the excess bandwidth is low, a fourth power nonlinearity . . . is better than the magnitude squared. In fact, fourth-power timing recovery
can even extract timing tones from signals with zero excess bandwidth. . . . Simulations
for QPSK . . . suggest that fourth-power circuits out-perform absolute-value circuits for
signals with less than about 20% excess bandwidth.

What about aliasing? Barry, Lee, and Messerschmitt continue:

If timing recovery is done in discrete-time, aliasing must be considered . . . Any nonlinearity will increase the bandwidth of the . . . signal . . . In the presence of sampling,
however, the high frequency components due to the nonlinearity can alias back into the
bandwidth of the bandpass lter, resulting in additional timing jitter. . . . Therefore, in
a discrete-time realization, a magnitude-squared nonlinearity usually has a considerable
advantage over either absolute-value or fourth-power nonlinearity.

What about attempting baud-timing recovery on a passband signal prior to full

downconversion to baseband? Barry, Lee, and Messerschmitt (2004) consider this
on page 750:
The same relative merits of squaring, absolute-value, and fourth-power techniques apply
to passband timing recovery as to baseband. In particular, absolute-value and fourthpower are usually better than squaring, except when aliasing is a problem in discretetime implementations. As with baseband signals, it is sometimes advantageous to prelter the signal before squaring.

Exercise 16.31. Use qamtimobj.m as a basis to calculate the ve objective

functions for 4-QAM using the SRRC pulse shape with rollo factors 0.1, 0.5,
and 0.7.
a. For each case, state whether the objective must be maximized or minimized.
b. Do any of the objective functions have local stationary points that are not
globally optimal? (Recall Figure 12.8, which shows local minima for PAM.)
c. Derive the corresponding adaptive element (i.e., the update equation for [k])
for one of the objective functions when using the SRRC. Which rollo factor
is best? Why?

384

Chapter 16. A Digital Quadrature Amplitude Modulation Radio

Exercise 16.32. Mimic clockrecOP.m on page 262 to implement the timingrecovery algorithm for [k] using objective function (e).
a. Derive the adaptive-element update for objective function (a).
b. Implement the adaptive element for objective (a).
c. Choose stepsizes so that the two methods converge in approximately the
same number of iterations.
d. Add noise to the two methods. Which of the methods is more resilient to the
added noise?
Exercise 16.33. State a dispersion-based objective function for timing recovery
of QAM analogous to that in Figure 12.11.
a. Plot the value of the objective function against the timing oset .
b. Derive the corresponding adaptive element, i.e., the update equation for [k].
c. Implement the adaptive element (mimicking clockrecOP.m on page 262).

16.7

Baseband Derotation
Carrier recovery identies the phase and frequency dierences between the modulating sinusoids at the transmitter and the demodulating sinusoids at the
receiver. Standard approaches such as the PLL or Costas loop of Section 16.4
operate at passband, that is, the iteration that identies the phases and frequencies operates at the sampling rate of the receiver, which is typically higher
than the symbol rate.
As observed in Section 16.3, a phase oset can be pictured as a rotation of
the constellation and a frequency oset can be interpreted as a spinning of the
constellation diagram. Of course these are baseband diagrams, which show
the symbol values after the timing recovery has downsampled the signal to the
symbol rate. This suggests that it is also possible to approach carrier recovery
at baseband, by constructing an adaptive element that can derotate a skewed
constellation diagram. This section shows how carefully designed adaptive elements can derotate a QAM constellation. These are essentially single-parameter
equalizers that operate on the two real data streams by treating the signal and
equalizer as complex-valued quantities.
Bingham (1988), on page 231, motivates the post-equalization carrier-recovery
process for QAM:
A complex equalizer . . . can compensate for any demodulating carrier phase, but it
is easier to deal with frequency oset by using a separate circuit or algorithm that,
because it deals with only one variable, carrier phase, can move faster without causing
jitter.

A Digital Quadrature Amplitude Modulation Radio

385

Figure 16.15 The complex-valued signal r(t) has

s(t)

e j

r(t)

e j

x(t)

been rotated through an (unknown) angle . If

this is passed through a complex factor ej with
, the complex-valued x(t) approaches the
unrotated message s(t).

The situation is this: some kind of (passband) carrier recovery has been done,
but it is imperfect and a slow residual rotation of the constellation remains.
Binghams rst sentence observes that a complex-valued equalizer (such as those
detailed in Section 16.8) can accomplish the derotation at the same time that it
equalizes, by multiplying all of the (complex) equalizer coecients by the same
complex factor ej . But equalizers have many parameters: real-valued equalizers
such as those in Chapter 13 can have dozens (and some real-world equalizers
have hundreds) of parameters. It takes longer to adjust a dozen parameters than
to adjust one. A dozen parameters introduce more jitter than a single parameter.
Even worse, consider a situation where a multitap equalizer has adapted to
perfectly equalize a complex channel. If the carrier now begins to develop a
rotational oset, all of the coecients of the equalizer will move (multiplied by
the factor ej ). Thus all of the otherwise well-set equalizer gains move in order
to accomplish a task that could be dealt with by a single parameter (the value
of the rotation ). Never send a dozen parameters to do a job when one will do!
To see how this works, model the (downsampled) received signal as r(t) =
ej s(t), where represents the carrier phase oset. A one-tap equalizer
is the system ej as shown in Figure 16.15. Clearly, if the parameter can
be adjusted to be equal to the unknown , the carrier oset is corrected and
x(t) = s(t).
As in the design of any adaptive element, it is necessary to choose an objective
function. There are several possibilities, and the remainder of this section details
two: a decision-directed strategy and a dispersion-minimization strategy. At the
end of the day, the derotated signals x1 and x2 will be sampled and quantized
to give the hard decisions s1 = sign(x1 ) and s2 = sign(x2 ), which can be written
more compactly as s = sign(x) with s = s1 + j
s2 .
Consider as objective function the average of
JBD () = (1/2)(s x)(s x) ,
where the superscript indicates complex conjugation. An element that adapts
so as to minimize this objective is

(s x)(s x) 
[k + 1] = [k]
.

2

=[k]
The objective can be rewritten
(s x)(s x) = (s1 x1 )2 + (s2 x2 )2
since the imaginary parts cancel out. Hence
x1
x2
JBD ()
= (s1 x1 )
(s2 x2 )
.

386

Chapter 16. A Digital Quadrature Amplitude Modulation Radio

The x can be rewritten explicitly as a function of since x = ej r, which is

x1 = r1 cos() r2 sin(),
x2 = r2 cos() + r1 sin()
when expressed in real and imaginary form (observe the similarity to Exercise
16.10). Accordingly,
x1
= r1 sin() r2 cos() = x2 ,

x2
= r2 sin() + r1 cos() = x1 .

Thus
JBD ()
= (s1 x1 )x2 (s2 x2 )x1 = s1 x2 s2 x1

leads to the adaptive update for 4-QAM

[k + 1] = [k] (s1 [k]x2 [k] s2 [k]x1 [k]),

(16.17)

which can be compared directly to Figure 16-6 in Barry, Lee, and Messerschmitt.
Implementation of (16.17) requires access to the signals s1 [k] and s2 [k]. This
can be accomplished using a training sequence, or more commonly by using a
decision-directed variation that replaces s1 [k] with s1 [k] and s2 [k] with s2 [k].
The adaptive derotator element can be examined in operation in
qamderotate.m, where the 4-QAM message is rotated by an angle -phi. After
the derotator, the x (and the hard decisions shat1 and shat2) restore the constellation to its proper orientation.
Listing 16.7. qamderotate.m derotate a complex-valued QAM signal
N=1000; j=sqrt ( 1 ) ;
% # symbols , time b a s e
s=pam (N, 2 , 1 ) + j * pam (N, 2 , 1 ) ;
% s i g na l s of length N
p h i = 0.1;
% a n g l e to r o t a t e
r=s . * exp( j * p h i ) ;
% rotation
mu= 0 . 0 1 ; t h e t a =0; t h e t a=zeros ( 1 ,N ) ;
% i n i t i a l i z e variables
f o r i =1:N1
% adapt th r o u g h a l l data
x=exp ( j * t h e t a ( i ) ) * r ( i ) ;
% x=r o t a t e d ( r )
x1=r e a l ( x ) ; x2=imag( x ) ;
% r e a l and i m a g i n a r y p a r t s
s h a t 1=sign ( x1 ) ; s h a t 2=sign ( x2 ) ;
% hard d e c i s i o n s
t h e t a ( i +1)= t h e t a ( i )mu* ( s h a t 1 * x2s h a t 2 * x1 ) ; % i t e r a t i o n
end

An alternative objective function based on dispersion minimization considers

the average of
JDM () = (Re{rej }2 )2 ,

(16.18)

where is a real constant. This objective function can be viewed as the projection
onto the real axis of a rotated, but otherwise perfect, symbol constellation. Since
the real parts of the 4-QAM symbols cluster around 1, choosing = 1 allows
JDM () to be zero when = . For other constellations, JDM ()|= will, on

A Digital Quadrature Amplitude Modulation Radio

387

average, be smaller than nearby values. The associated dispersion-minimizing

adaptive element is
[k + 1] = [k] (Re{rej }2 )Re{rej[k] }Im{rej[k] }.

(16.19)

Exercise 16.34. Rewrite the system in Figure 16.15 in its real (implementable)
form by letting s(t) = s1 (t) + js2 (t) and ejz = cos(z) + j sin(z). Express x(t)
directly as a function of the real-valued signals s1 (t) and s2 (t). Show that =
implies x(t) = s(t).
Exercise 16.35. Reimplement the qamderotate.m routine without using any
complex numbers.
Exercise 16.36. Calculate the derivative of JDM () of (16.18) with respect to
, which is used to implement the algorithm (16.19).
Exercise 16.37. Implement the dispersion-minimization derotator by mimicking qamderotate.m.
Exercise 16.38. Formulate a DD derotator for use with 16-QAM. Implement
it in imitation of qamderotate.m. Is the convergence faster or slower than with
4-QAM?
Exercise 16.39. Formulate a dispersion-minimization derotator for use with
16-QAM.
Exercise 16.40. Formulate a derotator for use with 16-QAM that uses only the
four corner points of the constellation, those with the highest SNR.

16.8

Equalization for QAM

Intersymbol interference (ISI) occurs when symbols interact, when the waveform
of one symbol overlaps and corrupts the waveform of other symbols nearby in
time. ISI may be caused by overlapping pulse shapes (as discussed in Chapter
11) or it may be caused by a nonunity transmission channel, physically a result of
multipath reections or frequency-selective dispersion. Chapter 13 showed that
for PAM, the (real-valued) received symbols can often be passed through a (realvalued) equalizer in order to undo the eects of the channel. It is possible to pass
both the real and the imaginary parts of a QAM signal through separate (realvalued) equalizers, but it is far more convenient to model the equalization portion
of a QAM receiver as complex-valued symbols passing through a complex-valued
equalizer.
This section shows that complex-valued equalizers are no more dicult to
design and implement than their real-valued counterparts. As in Chapter 13, the

388

Chapter 16. A Digital Quadrature Amplitude Modulation Radio

operation of the equalizer can be pictured in the frequency domain. For example,
Figures 13.10, 13.12, and 13.14 (on pages 296299) show frequency responses of
a channel and its converged equalizer. The product of the two is near unity:
that is, the product has magnitude response (approximately) equal to unity at
all frequencies. This is the sense in which the equalizer undoes the eect of the
channel. With complex-valued QAM, the signal has a frequency response that
is not necessarily symmetric about zero. The equalizer can thus be pictured
in exactly the same way (with the goal of attaining a unity product) but the
frequency response of the equalizer may also need to be asymmetric in order to
achieve this goal. Thus the equalizer must be complex-valued in order to achieve
an asymmetric frequency response.
The rst step in the design of any adaptive element is to choose an objective
function. One approach, analogous to the LMS equalizer of Section 13.3, is to
presume that a (complex-valued) training sequence d[k] is available. The output
of the equalizer can be written as an inner product of a set of n + 1 (complexvalued) lter coecients fi with a vector of (complex-valued) received signals
r[k i] as

f1
n
f2


y[k] =
fi r[k i] = (r[k], r[k 1], . . . , r[k n])T . = X T [k]f. (16.20)
..
i=1

fn
Let e[k] be the dierence between the training signal d and the output of the
equalizer
e[k] = d[k] X T [k]f,

(16.21)

and dene the objective function

1
avg{e[k]e [k]}.
2
From now on, everything in this section is complex-valued, and any term can be
broken into its real and imaginary parts with the appropriate subscripts. Hence
JLMS (f ) =

e[k] = eR [k] + jeI [k],

d[k] = dR [k] + jdI [k],
X[k] = XR [k] + jXI [k],
f [k] = fR [k] + jfI [k].
The lter coecients f can be updated by minimizing JLMS (f ) using
#








,
[e[k]e [k]]
+j
[e[k]e [k]]
f [k + 1] = f [k]

fR
fI
fR =fR [k]
fI =fI [k]

A Digital Quadrature Amplitude Modulation Radio

389

where indicates complex conjugation. The output of the equalizer y[k] =

X T [k]f can be expanded as
T
[k] + jXIT [k])(fR + jfI )
X T [k]f = (XR
T
T
[k]fR XIT [k]fI ) + j(XIT [k]fR + XR
[k]fI ),
= (XR

which has complex conjugate

T
T
(X T [k]f ) = (XR
[k]fR XIT [k]fI ) j(XIT [k]fR + XR
[k]fI ).

Note that
e[k]e [k] = e2R [k] + e2I [k],
which can be expanded using (16.21) as
T
eR [k] = Re{e[k]} = Re{d[k]} Re{X T [k]f } = dR [k] XR
[k]fR + XIT [k]fI ,
T
[k]fI .
eI [k] = Im{e[k]} = Im{d[k]} Im{X T [k]f } = dI [k] XIT [k]fR XR

Since the training sequence d is independent of the lter f , the derivative

d/f = 0, and
e[k]e [k]
e2 [k] + e2I [k]
= R
fR
fR
2
e [k] eR [k] e2I [k] eI [k]
= R
+
eR [k] fR
eI [k] fR
= 2eR [k]XR [k] 2eI [k]XI [k].
Similarly,
e[k]e [k]
= 2eR [k]XI [k] 2eI [k]XR [k].
fI
Therefore the update is
f [k + 1] = f [k] +
(eR [k]XR [k] + eI [k]XI [k] + j(eR [k]XI [k] + eI [k]XR [k])),
which simplies to
f [k + 1] = f [k] + e[k]X [k].

(16.22)

Compare this with the real version of trained LMS for PAM in (13.27). Except
for the notation (the present version is written with vectors), the only dierence
is the complex conjugation.
The complex-valued 4-QAM equalizer LMSequalizerQAM.m passes a 4-QAM
signal through an arbitrary (complex) channel to simulate the occurrence of ISI.
The equalizer f adapts, and, with luck, converges to a lter that is roughly
the inverse of the channel. The equalizer can be tested by generating a new
set of data ltered through the same channel and then ltered through the
equalizer. If it is working, the doubly ltered signal should be almost equal to
the original, though it will likely be shifted in time as it passes through the two

390

Chapter 16. A Digital Quadrature Amplitude Modulation Radio

lters. Observe that, in Matlab, when the transpose operator is applied to

a complex-valued vector, the result is the complex-conjugated transpose. Hence
the nal line directly implements (16.22).
Listing 16.8. LMSequalizerQAM.m 4-QAM LMS equalizer
b = [ 0 . 5 + 0 . 2 * j 1 0.6 0.2 * j ] ;
m=1000;
s=pam (m, 2 , 1 ) + j * pam (m, 2 , 1 ) ;
r=f i l t e r ( b , 1 , s ) ;
n=4; f=zeros ( n , 1 ) ;
mu= . 0 1 ; d e l t a =2;
y=zeros ( n , 1 ) ;
f o r i=n+1:m
y ( i )= r ( i : 1 : i n+1) * f ;
e=s ( i d e l t a )y ( i ) ;
f=f+mu* e * r ( i : 1 : i n + 1 ) ;
end

%
%
%
%
%
%
%
%
%
%
%

d e f i n e complex c h a n n e l
how many data p o i n t s
4QAM s o u r c e o f l e n g t h m
output o f channel
i n i t i a l i z e e q u a l i z e r at 0
s t e p s i z e and d e l a y d e l t a
p l ace to s t o r e output
iterate
output o f e q u a l i z e r
c a l c u l a t e e r r o r term
update e q u a l i z e r c o e f f i c i e n t s

There are several ways to test how well the equalizer is working. One simple
way is to look at the convolution of the channel with the nal converged equalizer;
if this is close to a pure (possibly multisample) delay (i.e., an impulse response
with a 1 in the ith location and zeros elsewhere), all is well. For example, with
the default channel b above, the converged equalizer f is approximately
0.31 0.019j, 0.66 0.15j, 0.34 + 0.01j, 0.14 + 0.03j.
and the convolution of b and f is
0.15 0.07j, 0.05 + 0.04j, 1.01 0.01j,
0.02 + 0.01j, 0.06 0.04j, 0.08 0.05j,
which is reasonably close to 0, 0, 1, 0, 0, 0.
The test routine LMSequalizerQAMTest.m uses the nal converged value f of
the equalizer and the denition of the channel b from LMSequalizerQAM.m. A
new set of data is generated, passed through the equalizer, and then quantized
to form the hard decisions dec. The vector err counts up the number of errors
that have occurred for each delay. If one of the entries is zero then the equalizer
has done its job perfectly for that delay. The loop in Listing 16.9 searches over
a number of dierent time shifts (the terms sh) because the output will be a
delayed version of the input and the best delay might not be known beforehand.
In practice, a typical equalizer uses only one delay.
Listing 16.9. LMSequalizerQAMTest.m tests nal equalizer lter
f i n a l e q=f ;
% test final f i l t e r f
m=1000;
% new data p o i n t s
s=pam (m, 2 , 1 ) + j * pam (m, 2 , 1 ) ;
% new 4QAM s o u r c e o f l e n g t h m
r=f i l t e r ( b , 1 , s ) ;
% output o f channel
y t=f i l t e r ( f , 1 , r ) ;
% u s e f i n a l f i l t e r f to t e s t
dec=sign ( r e a l ( y t ))+ j * sign ( imag( y t ) ) ; % q u a n t i z a t i o n
f o r sh =0: n
% i f e q u a l i z e r working , one
e r r ( sh +1)=0.5 *sum( abs ( dec ( sh +1:end) s ( 1 : endsh ) ) ) ;
end
% of these delays = zero error

A Digital Quadrature Amplitude Modulation Radio

391

Exercise 16.41. Derive a decision-directed equalizer for 4-QAM analogous to

the DD equalizer for PAM in Section 13.4. Implement the DD equalizer starting
with LMSequalizerQAM.m. Find a channel b for which the LMS equalizer provides
a correct equalizer while the DD equalizer does not.
Exercise 16.42. Another way of investigating the quality of an equalizer is to
plot the frequency responses. Use freqz to plot the frequency response of the
default channel and of the resulting equalizer. How close is their product to a
transfer function that has magnitude unity at all frequencies? Because the signals
are complex, it is necessary to plot both positive and negative frequencies. The
phase spectrum should be approximately a straight line. Can you relate the slope
of this line to the total delay in the lter?
Exercise 16.43. Show that a complex dispersion-minimization equalizer for 4QAM (analogous to the dispersion-minimization equalizer for PAM in Section
13.5) is given by the update
f [k + 1] = f [k] + X[k]y [k]( |y[k]|2 ),
where y[k] is dened in (16.20) and where = 1. Implement the equalizer starting with LMSequalizerQAM.m. Discuss the advantages and disadvantages of this
equalizer in comparison with LMS.
Exercise 16.44. Modify LMSequalizerQAM.m to generate a source sequence
from the 16-QAM alphabet a jb, where a and b can be be 1 or 3. For the
default channel [0.5+0.2*j 1 -0.6-0.2*j], nd an equalizer that opens the
eye.
a. What equalizer length n is needed?
b. What delays delta give zero error in the output of the quantizer?
c. Is this a fundamentally easier or more dicult task than equalizing a 4-QAM
signal?
Exercise 16.45. Derive and implement a dispersion-minimization equalizer for
16-QAM using the update in Exercise 16.43. Hint: the 4-QAM algorithm is still
OK but = 1 might scale the constellation improperly.

16.9

Alternative Receiver Architectures for QAM

There are many ways to organize a digital receiver. This section takes a glance
at the QAM literature and shows how the components of the receiver can be
shued and rearranged in a surprising number of ways. Bingham (1988) calls the
receiver in Figure 16.16 the analog receiver because the ltering, downconversion,

Chapter 16. A Digital Quadrature Amplitude Modulation Radio

and timing recovery are done before the sampling into digital form. The decision
blocks quantize to the points of the QAM constellation.
The most interesting feature of Binghams second receiver, shown in Figure
16.17, is the decision-directed carrier recovery. The outputs of the quantizer are
assumed to be correct and are fed back to adjust the phase of the sinusoids in
the mixers, as discussed in Section 16.4.2.
The third of Binghams receiver structures is the nearly-all-digital receiver
of Figure 16.18. This uses a free-running sampler and digital interpolation to estimate the value of the signals at the desired sampling times, as is familiar from
Chapter 12. There is a free-running downconversion before the equalizer and
post-equalization carrier recovery (as described in Section 16.4.2) that operates
on the hard decisions. The equalization element is also decision-directed. Interestingly, the AGC also incorporates feedback from the same signals entering the
carrier recovery.
The nal architecture from Bingham is the all-digital receiver of Figure
16.19. The initial sampling is accomplished by a free-running sampler that operates at a sub-Nyquist rate (compared with the RF of the carrier). Again, the
AGC incorporates feedback on the magnitude of the error between the output of
the equalizer and the hard decisions. The control of the timing recovery is purely
digital and requires an interpolator. The downconversion is accomplished in two
stages: a xed free-running oscillator followed by a decision-directed derotation.
The decision-feedback equalizer is a nite-impulse-response lter that operates
in the feedback loop. It is adapted using a decision-directed objective function.

BPF

Carrier
Recovery

cos(2f0t)

Timing
Recovery

sin(2f0t)

Decoder

LPF

LPF
Figure 16.16 Binghams analog receiver for QAM carries out the bulk of the processing

in analog circuitry before sampling. This is a redrawing of Figure 5.1 in Bingham

c
(1988). The gure is 1988
John Wiley & Sons, Inc., and is reprinted with
permission.

LPF
BPF

AGC

cos(2f0t)
sin(2f0t)

DD Carrier
Recovery

LPF
Timing
Recovery

Decoder

392

Figure 16.17 The second

of Binghams QAM
radio designs uses a
decision-directed carrier
recovery. This is a
redrawing of Figure 5.2
in Bingham (1988). The
c
gure is 1988
John
Wiley & Sons, Inc., and
is reprinted with
permission.

393

A Digital Quadrature Amplitude Modulation Radio

Magnitude
Error
interpolation
filter
Phase
AGC ADC
BPF Splitter LPF
Timing
Recovery

Rotator

Equalizer
e j

e j2fckTs

Carrier
Recovery

e j

Derotator
Figure 16.18 Binghams third receiver digitizes near the start of the signal chain and

adapts many of its elements using feedback from the hard decisions. Redrawn from
c
Figure 5.3 of Bingham (1988). The gure is 1988
John Wiley & Sons, Inc., and is
reprinted with permission.

ADC

BPF

Phase
AGC Splitter

Interpolator
Timing
Recovery

Allpass
Filter
e j2fckTs

e j
e j

Equalizer
+

Carrier
Recovery

Magnitude
Error
Figure 16.19 Binghams all-digital receiver structure is a redrawing of Figure 5.4 of
Bingham (1988). The receiver samples at the start of the signal chain and carries out
c
all processes in the digital domain. The gure is 1988
John Wiley & Sons, Inc., and
is reprinted with permission.

The receiver structures in Figures 16.16 to 16.19 follow the development of the
generic receivers in Figure 12.2 on page 252 which progress from mostly analog
to completely digital. Many of the details in the structures can be rearranged
without undue concern: for instance, a decision-feedback equalizer could replace
a feedforward equalizer, the adaptive elements could operate by optimizing a dispersion objective rather than a decision-directed objective, etc. The possibilities
are staggering!
A typical digital QAM receiver is presented in Figure 4-2 of Meyr,
Moeneclaey, and Fechtel (1998), and is shown here in Figure 16.20. The front
end uses an analog mixer for coarse downconversion and a digital derotation (in
the block labeled phase estimator) fed back from after the downsampler. The
free-running sampling is done at a rate Ts faster than the symbol rate T , and
uses digital interpolation from the timing recovery. Observe that the matched
lter occurs after the interpolation but before the timing element. This helps to

394

Chapter 16. A Digital Quadrature Amplitude Modulation Radio

f=1/Ts
r(t)

Analog
Prefilter

ej(0+)t

rf (t)

rf (kTs)

unTs

Interpolator
Filter

Frequency
Error
Detection

mn
z(nT + T)

Matched
Prefilter

Symbol an
Detection

mnTs
e j(mnTs)

Frequency
Integrator
Frequency
Synthesizer

Timing
Recovery

Phase
Estimator
Frequency
Integrator

Coarse
Frequency
Correction
Figure 16.20 This typical digital QAM receiver is adopted from Figure 4-2 in Meyr,

Moeneclaey, and Fechtel (1998). Two sampling rates are shown, the bulk of the
processing uses a sampling period Ts that is faster than the symbol rate T . The gure
c
is 1998
John Wiley & Sons, Inc., and is reprinted with permission.

smooth the signal for the timing recovery, which also feeds forward to control
the downsampling to the symbol rate.
Figure 16.21 shows a generic equalized data demodulator from Treichler,
Larimore, and Harp (1998). The most prominent feature of this design is the
control-loop architecture, which emphasizes how tightly coupled the various components are. As in previous structures, the mixing is accomplished by a local
oscillator that has feedback from the output of the decision device. A second
local oscillator down-modulates again after the equalizer, which is working faster
than the symbol rate. Two equalizers are shown, a feedforward equalizer and a
decision-feedback equalizer.
The generic blind demodulator for QAM shown in Figure 16.22 is also taken
from Treichler, Larimore, and Harp (1998). The control sequencer amalgamates signals from various places in the receiver and uses them to create a more
centralized control structure, embedding various elements within the loops of others. As before, the analog AGC uses feedback correction from the free-running
sampler output. The quadrature free-running downconversion is trimmed by the
hard decisions and the resampler uses a timing-recovery method that adapts
using only the largest values in the constellation. The objective function of the
equalizer is selected dynamically from either LMS (with a training sequence) or
CMA, which uses the dispersion minimization for blind adaptation.
The multipoint network modem of Jablon (1992) is shown in Figure 16.23.
The design uses a sampler with an adjustable clock rate, a passband equalizer
with de-spun error, and a post-equalization carrier recovery.
The nal architecture of this section is Lee and Messerschmitts typical passband QAM receiver that begins with an analog bandpass lter and uses a

395

A Digital Quadrature Amplitude Modulation Radio

Soft
Decisions

Forward
Error
Correction

Hard
Decisions
(w/ FEC)

DFE
BPF

AGC

Hard
Decisions
(no FEC)

Equalizer

Timing
Recovery

Power
Estimation

Equalizer
Adjustment

Carrier
Recovery

Control-Loop Bus
Figure 16.21 This generic equalized data demodulator, redrawn from Figure 4 in

Treichler, Larimore, and Harp (1998), emphasizes the degree of coordination in the
processing of the receiver. Signals are fed back from the output of the hard decisions
and from the decision-directed carrier recovery. Many of the loops include signicant
(and complicated) portions of the receiver such as the feedforward equalizer and the
decision-feedback equalizer. Such nesting of adaptive elements is a tricky business
c
since they must must not interfere with each other. The gure is 1998
IEEE and is
used with permission.
Soft
Decisions
BPF

AGC

A/D

QDC

Hard
Decisions

Equalizer
CMA

Power
Estimation

Timing
Recovery

Equalizer
Adjustment

LMS

Carrier
Recovery

Control Sequencer
Figure 16.22 This receiver can switch between two equalizers: LMS for faster

convergence when there is a training sequence available and CMA (dispersion

minimization) when there is no training. Redrawn from Figure 11 in Treichler,
c
Larimore, and Harp (1998). The gure is 1998
IEEE and is used with permission.

sampler adjusted by a decision-directed method as shown in Figure 16.24. The

phase splitter is followed by a preliminary free-running downconversion and then
downsampled to T /2 sampling, again with a decision-directed adjustment. The
fractionally spaced equalizer has a de-spun decision-directed adjustment and
the digital downconversion uses a decision-directed carrier recovery. The decisionfeedback equalizer at the end is also decision-directed.

396

Chapter 16. A Digital Quadrature Amplitude Modulation Radio

A/D

AGC

Decoding

Equalizer
PLL

Sampling
Clock

Timing
Recovery

e j[k]

e j[k]

Error
Computation
Figure 16.23 This QAM receiver, simplied from the original Figure 2 in Jablon

(1992), uses a passband equalizer before the demodulation and trimming by the PLL.
c
The gure is 1992
IEEE and is used with permission.

Timing Recovery
Nyquist-rate
sampling

BPF

Twice
symbol-rate
sampling

Symbol-rate
sampling

Fractionally
Spaced
Precursor
Equalizer

Phase
Splitter
e j2fckTs
Preliminary
demodulation

Carrier
Recovery

Hard
Decisions

+
+

Postcursor
Equalizer
Figure 16.24 This typical passband QAM receiver is redrawn from Figure 5-21 of

c
Barry, Lee, and Messerschmitt (2004). The gure is 2004
Kluwer Academic, and is
used with the kind permission of Springer Science and Business Media.

Why are there so many receiver architectures? Why not gure out once and
for all which is the best, and stick with it? One reason is that some structures are
more suited to particular operating circumstances than others: some components,
some objective functions, and some ordering of the elements may be better in
certain situations. For example, Barry, Lee, and Messerschmitt (2004) state the
following on page 733:
One practical diculty arises when an adaptive equalizer is used in conjunction with a
decision-directed carrier recovery loop. Baseband adaptive equalizers assume that the
input has been demodulated. The solution to this diculty . . . is to use a passband
equalizer. . . . By placing the forward equalizer before the carrier recovery demodulation,
we avoid having the equalizer inside the carrier recovery loop. By contrast, a baseband
equalizer would follow the demodulator and precede the slicer. This means that it is
inside the carrier recovery loop. Consequently, the loop transfer function of the carrier
recovery includes the time-varying equalizer, causing considerable complication. At
the very least, the long delay (several symbol intervals) associated with the baseband
equalizer would force the loop gain of the carrier recovery to be reduced to ensure
stability, impairing its ability to track rapidly varying carrier phase. The passband
equalizer . . . mitigates this problem by equalizing prior to demodulation.

A Digital Quadrature Amplitude Modulation Radio

397

From page 429 of Gitlin, Hayes, and Weinstein (1992):

At low SNR, the designer can of course use the nonlinear/PLL carrier recovery scheme
. . . but at moderate-to-high SNR levels, when data decisions reliably replicate the
transmitted data, the data-directed loop has become the preferred carrier recovery
system.

These represent only a handful of the many possible architectures for a QAM
radio. The next section asks you to explore some alternatives.

The Q3 AM Prototype Receiver

16.10

Chapter 15 asked you to code the complete M6 receiver using only the tiny
building blocks from the homework and book chapters. The task with the QuasiQuixotic QAM receiver (Q3 AM for short) is somewhat dierent. A complete
(working) prototype was coded by a crack team of students at Cornell University
in winter of 2002.1 The routine qpskRx.m can be found on the website, and the
transmission protocol and receiver architecture are described in the next two
sections. Your job, should you choose to accept it, is to make changes to this
receiver: to improve the radio by using better parts (such as a PLL in place of
a Costas loop), to explore alternative structures (such as those of the previous
section), and to adapt it to work with bigger constellations (such as 16-QAM).
As Section 16.9 showed, there are many ways to structure a QAM receiver.
Exploring various rearrangements of the modules and submodules should provide
insight into the inner working of the radio at a slightly higher level than when
designing the M6 radio. There are a variety of disturbances and impairments
that the Q3 AM receiver will be asked to handle, including

r
r
r
r
r
r
r
r

a carrier frequency that may dier from its nominal value

a clock frequency that may dier from its nominal value
phase noise that may vary over time
a timing oset that may vary over time
multipath interference
time-varying multipath gains
wideband additive channel noise
interference from other narrowband signals nearby in frequency

These impairments also bedeviled the M6 design challenge.

The receiver will include many xes for these impairments:
1. a free-running four times oversampled (relative to baud interval) received
passband signal
2. a mixer with phase adaptation via dual quadriphase Costas loop
1

Many thanks to J. M. Walsh, C. E. Orlicki, A. G. Pierce, J. W. Smith, and S. R. Leventhal.

398

Chapter 16. A Digital Quadrature Amplitude Modulation Radio

3. lowpass ltering for downconversion, matched lter, and interpolation all provided by a matched lter with adjusted timing oset adapted with maximization of the fourth power of downsampled signals in dual-loop conguration
4. correlation used to resolve phase ambiguity and to locate the start of the
training sequence in the equalizer input
5. linear equalizer adaptation via LMS; switched to decision-directed LMS adaptation during the data (i.e. non-training) portion
6. frame-synchronized descrambler and (5, 2) linear block-code decoder

16.11

Q3 AM Prototype Receiver Users Manual

This section outlines the operation and usage of the Q3 AM software radio,
including the specications for the transmitter and the receiver. The full structure is diagrammed in Figure 16.25 and the .m software routines may be found
on the website.

16.11.1

Transmitter and Impairments Generated in the Received Signal

The transmitter and channel portions of the software use a number of parameters
as shown in Tables 16.1 and 16.2 in order to generate a sampled received signal
with the following impairments.

Oversampled
Carrier Synchronized
Signal
Carrier Recovery

Baud-Spaced
Clock-Synchronized
Signal

Timing Recovery

training
data

Trained LMS
Equalization
initial
equalizer

Correlation

Oversampled
Output from
Analog Front End

Decision-Directed
Equalization
Equalized and
Quantized Message
Comparison
and BER calculation

Original
Message
Receiver Inputs

Bit Error Rate

Recovered
Message

Descrambling
and Decoding
Receiver Output

Figure 16.25 An overview of the receiver simulator and tester.

A Digital Quadrature Amplitude Modulation Radio

399

Phase Noise: The carrier phase noise is modeled as a random walk, such that
k+1 = k + k ,

(16.23)

where k is a white noise whose variance is dictated when running the software.
(See Section 16.11.3.)
Timing Oset Noise: The timing oset noise is modeled in the same manner
as the phase noise, only with a white noise of a dierent variance, which is
specied when running the software.
Carrier Oscillator Frequency and Phase Oset: The frequency oset between
the transmitter and receiver carrier oscillators is specied as a ratio, of f ,
such that fact = (1 of f )fc . The phase oset can be specied in any region
from 0 to 2.
Baud Oscillator Frequency and Phase Oset: The frequency oset between the
transmitter and receiver baud oscillators (aka clocks) is also specied as a
ratio, of f , such that Tsact = Ts of f /(1 of f ). The baud oscillator phase
oset, which is also known as the xed timing oset, may also be specied.
Multipath Channels: The software has the ability to simulate transmission
through a xed multipath channel specied at run time.
Time-Varying Multipath: It is possible to study the receivers ability to track
timing-varying multipath by turning on a ag in the software and specifying
the speed of time variation for a particular built-in time-varying multipath
channel. (See Section 16.11.3.)
Wideband Additive Channel Noise The received signal includes additive noise,
which is modeled as a white Gaussian random process. This noise is added in
the channel, and its variance can be specied by specifying the channel SNR
when running the software.

16.11.2

The Receiver
Carrier Recovery and Downconversion
The input to the receiver from the channel and transmitter is an upconverted
signal centered in frequency at fc , and oversampled by a factor of four. The
rst block this data enters is the carrier-recovery unit, which uses a quadriphase
Costas loop. The estimates, est [k], from the adaptive Costas loop are then used
to demodulate the signal by multiplication by exp(j(2fc Ts k + est [k]). The
lowpass ltering done in the timing-recovery unit by the matched lter completes
the demodulation.

Timing Recovery
The oversampled signal is then sent to the baud-timing-recovery unit, which
interpolates the signal using a pulse shape with nonzero timing oset (with the
timing oset estimated adaptively) matched to the transmitter pulse shape. This
is slightly dierent from sinc-interpolation, because it uses the time-shifted pulse

400

Chapter 16. A Digital Quadrature Amplitude Modulation Radio

Table 16.1. The parameters which must be set in the rst structure, b, passed into the
transmitter portion of the software suite. See the example in Section 16.11.3 for more
details.

Name
debugFlag
ph off
f off
t off
SNR
channel
tNoiseVar
bfo
pNoiseVar
TVChan
alpha
c
g

Description
Flag to indicate whether or not to generate plots
Phase oset between the transmitter and receiver carrier
Frequency oset between the transmitter and receiver carrier
Timing oset between the transmitter and receiver baud
Signal-to-noise ratio in the channel
Fixed channel impulse response (ignored if TVChan = 1)
Variance of noise in the receiver baud timing
Frequency oset between the transmitter and receiver baud
Variance of phase noise in receiver and transmitter carrier
= 0, use static channel dened by channel parameter
= 1, use [1, 0, 0, 0, c + g cos(2fc kTs )]
Controls parameter variation in time-varying channel
Mean value c of the time-varying channel tap
Variation g about mean value of time-varying channel tap

Table 16.2. The parameters which must be specied in the second structure, specs, when
running the transmitter portion of the the software suite. See the example in Section 16.11.3
for more details.

Name
v
codingF
scrambleF1
beta
scrambleF2
OverSamp

Description
Square QAM constellation has 22v points (v = 1 for QPSK)
Whether or not to use coding (same as for transmitter)
Whether or not to generate diagnostic plots when scrambling
The rollo factor used in the SRRC matched lter
Whether or not to scramble (same as for transmitter)
Oversampling factor, usually set to 4

shape to interpolate instead. The timing oset is estimated using an adaptive

fourth-power maximization scheme. The signal is then downsampled to symbolspaced samples.

Correlation
Correlation must be performed to nd the segment used to train the equalizer,
thus this block correlates the signal with the pre-arranged training data. Once
the training segment is found, the correlation block passes the starting location
on to the equalizer along with the training data.

Equalization
Equalization is performed rst using an adaptive trained LMS algorithm, and
then, when the training runs out, using the adaptive decision-directed LMS

A Digital Quadrature Amplitude Modulation Radio

401

algorithm. Step sizes for this algorithm (along with those mentioned above) may
be specied as inputs to the software.

Descrambling and Decoding

In order to lower the probability of symbol error, the data are mapped to binary
and coded at the transmitter using a (5, 2) linear block code. Because coding correlates the data, scrambling with a length-21 random sequence is then performed
to restore signal whiteness. Scrambling is done by exclusive-oring the data
with a length-21 sequence. Descrambling is then performed at the receiver by
exclusive-oring with the same sequence. The synchronization for the descrambling is aided by the equalizer training data. Just as soon as the equalizer training
data end, the descrambling begins.

Performance Evaluation
After the receiver has nished processing the received signal, its soft decisions
(data at the equalizer output) are compared with the original sent symbols
in order to determine the mean-squared error at the equalizer output. It is
then assumed that the dispersion around the decisions is due to additive circularly Gaussian noise, and a probability of symbol error is determined using
this assumption and the average of the squared error over all data after the
training phase. Also, the hard decisions on the equalized, pre-decoding data
are compared with the original coded message symbols to give an empirical
SER. Furthermore, the hard decisions are descrambled and decoded, and the bit
error rate is measured by comparing the transmitted (uncoded) data. Thus, an
uncoded SER based on the mean-squared error, a measured pre-decoder SER,
and an empirical coded BER are determined. If the coded BER is suciently low
and the errors are sparsely distributed so that there is no more than one error
per symbol, the coded SER can be approximated by multiplying the BER (units:
errors/bit) by the number of bits/symbol (i.e., 2 for 4-QAM). When there is only
a small amount of data running through the system, the rst method with its
MSE-based approximation will be capable of more accurately estimating smaller
probabilities of error, because the MSE becomes substantial before the SER does.

16.11.3

Using the Software

The parameters that must be specied when running the software are all listed
and explained in Table 16.3. These parameters must be within a structure that
is passed to the software suite as an argument. An example run can be found
in sampleTest.m. In this example, m and specs were the structures that were
used to pass the arguments into the transmitter software, b was the structure
used to pass parameters into the receiver software, t was the SER derived from a
Gaussian approximation of the noise, z was the measured pre-decoding SER,
and c was the measured BER after decoding.

Chapter 16. A Digital Quadrature Amplitude Modulation Radio

Table 16.3. The parameters to specify when running the receiver portion of the software
suite. These parameters are all within a structure that is passed to the receiver port of the
software when running a simulation. See the example of Section 16.11.3.

Name
Costas Mu1
Costas Mu2
Time mu1
Time mu2
EqMu
ddEqMu
debugFlag
scrambleF
codingFlag
OverSamp

Description
First step size to use in the type-II Costas loop
Second step size to use in the type-II Costas loop
First step size to use in the dual timing recovery
Second step size to use in the dual timing recovery
Step size to use with the trained LMS equalizer
Step size to use with the decision-directed equalizer
Flag that indicates whether or not to generate plots
(If debugFlag = 1, the plots are drawn)
Flag that turns on and o the scrambling of the signal
Flag that turns on and o the coding of the signal
Oversampling factor, usually set to 4

The code to run a performance study that shows how the symbol error rate of
the scheme varies with the SNR in the channel, as plotted in Figure 16.26, can
be found in snrTest.m.
10

10

SER as the SNR increases

SER

402

10

10

Figure 16.26 The plot generated

by the example performance
study code. snrTest

Approximate
Measured

10

15

20

SNR

References
The following papers and books were utilized in preparing this chapter and are
recommended as more comprehensive sources on QAM radio.
J. B. Anderson, Digital Transmission Engineering, Prentice-Hall, 1999.
J. R. Barry, E. A. Lee, and D. G. Messerschmitt, Digital Communication, 3rd
edition, Kluwer Academic, 2004.
J. A. C. Bingham, The Theory and Practice of Modem Design, Wiley, 1988.

A Digital Quadrature Amplitude Modulation Radio

403

R. D. Gitlin, J. F. Hayes, and S. B. Weinstein, Data Communication Principles,

Plenum Press, 1992.
N. K. Jablon, Joint Blind Equalization, Carrier Recovery, and Timing Recovery
for High-Order QAM Signal Constellations, IEEE Transactions on Signal
Processing, vol. 40, no. 6, pp. 13831398, June 1992.
H. Meyr, M. Moeneclaey, and S. A. Fechtel, Digital Communication Receivers:
Synchronization, Channel Estimation, and Signal Processing, Wiley, 1998.
J. G. Proakis and M. Salehi, Communication Systems Engineering, 2nd edition,
Prentice-Hall, 2002.
J. R. Treichler, M. R. Larimore, and J. C. Harp, Practical Blind Demodulators
for High-Order QAM Signals, Proceedings of the IEEE, vol. 86, no. 10, pp.
19071926, October 1998.

Transforms, Identities, and

Formulas

This appendix gathers together all of the math facts used in the text. They are
divided into six categories:

r
r
r
r
r
r

Trigonometric identities
Fourier transforms and properties
Energy and power
Z-transforms and properties
Integral and derivative formulas
Matrix algebra

So, with no motivation or interpretation, just labels, here they are.

A.1

Trigonometric Identities
r Eulers relation
ejx = cos(x) j sin(x)

(A.1)

r Exponential denition of a cosine

1 jx
e + ejx
2

(A.2)

sin(x) =

1 jx
e ejx
2j

(A.3)

cos2 (x) =

1
(1 + cos(2x))
2

(A.4)

sin2 (x) =

1
(1 cos(2x))
2

(A.5)

cos(x) =

r Exponential denition of a sine

r Cosine squared
r Sine squared

Transforms, Identities, and Formulas

405

r Sine and cosine as phase shifts of each other




x = cos x

 2
 2
x = sin x
cos(x) = sin
2
2
Sinecosine product
sin(x) = cos

(A.8)

1
[cos(x y) + cos(x + y)]
2

(A.9)

1
[cos(x y) cos(x + y)]
2

(A.10)

r Cosinecosine product
cos(x) cos(y) =

r Sinesine product
r Odd symmetry of the sine

(A.7)

1
[sin(x y) + sin(x + y)]
2

sin(x) cos(y) =

sin(x) sin(y) =

(A.6)

sin(x) = sin(x)

(A.11)

r Even symmetry of the cosine

cos(x) = cos(x)

(A.12)

cos(x y) = cos(x) cos(y) sin(x) sin(y)

(A.13)

sin(x y) = sin(x) cos(y) cos(x) sin(y)

(A.14)

r Cosine angle sum

r Sine angle sum

A.2

Fourier Transforms and Properties

r Denition of Fourier transform


W (f ) =

w(t)ej2f t dt

(A.15)

r Denition of inverse Fourier transform



w(t) =

r Fourier transform of a sine


A  j
e (f f0 ) + ej (f + f0 )
2
Fourier transform of a cosine

A  j
e (f f0 ) + ej (f + f0 )
F {A cos(2f0 t + )} =
2
F {A sin(2f0 t + )} = j

W (f )ej2f t df

(A.16)

(A.17)

(A.18)

406

Appendix A. Transforms, Identities, and Formulas

r Fourier transform of impulse

F {(t)} = 1

(A.19)

r Fourier transform of rectangular pulse

With


(t) =

T /2 t T /2
otherwise,

(A.20)

sin(f T )
T sinc(f T ).
f T

(A.21)

1
0

F {(t)} = T

r Fourier transform of sinc function

F {sinc(2W t)} =
2W

f
2W


(A.22)

r Fourier transform of raised cosine

With


w(t) = 2f0

sin(2f0 t)
2f0 t


cos(2f t)
,
1 (4f t)2




F {w(t)} = 1 1 + cos (|f | f1 )

2
2f

(A.23)

|f | < f1
f1 < |f | < B

(A.24)

|f | > B,

where the rollo factor is = f /f0 .

r Fourier transform of square-root raised cosine (SRRC)

With w(t) given by
1 sin((1 )t/T ) + (4t/T ) cos((1 + )t/T )

t = 0, (T /4)
(t/T )(1 (4t/T )2 )
T
1
(1 + (4/))
t=0
T


  

 
+2

sin
+
cos
t = T /(4),

4
2T

1/2

 
1
(|f | f1 )
F {w(t)} =
1 + cos

2f

2
0

|f | < f1
f1 < |f | < B
|f | > B.

r Fourier transform of periodic impulse sampled signal

With
F {w(t)} = W (f )

(A.25)

(A.26)

Transforms, Identities, and Formulas

407

and

ws (t) = w(t)
F {ws (t)} =

1
Ts

r Fourier transform of a step

With

(t kTs ),

k=

W (f (n/Ts )).

(A.27)

(A.28)

n=

A t>0
0
t < 0,


1
(f )
+
.
F {w(t)} = A
2
j2f
w(t) =

(A.29)

r Fourier transform of ideal /2 phase shifter (Hilbert transformer)

lter impulse response
With


w(t) =


F {w(t)} =

1/(t)
0

t>0
t < 0,

j
j

f > 0,
f < 0.

(A.30)

r Linearity property
With F {wi (t)} = Wi (f ),
F {aw1 (t) + bw2 (t)} = aW1 (f ) + bW2 (f ).

(A.31)

r Duality property
With F {w(t)} = W (f ),
F {W (t)} = w(f ).

(A.32)

r Cosine modulation frequency-shift property

With F {w(t)} = W (f ),

1  j
e W (f fc ) + ej W (f + fc ) .
2
Exponential modulation frequency-shift property
With F {w(t)} = W (f ),
F {w(t) cos(2fc t + )} =

F {w(t)ej2f0 t } = W (f f0 ).

(A.33)

(A.34)

r Complex conjugation (symmetry) property

If w(t) is real-valued,
W (f ) = W (f ),

(A.35)

where the superscript denotes complex conjugation (i.e., (a + jb) = a jb).

In particular, |W (f )| is even and  W (f ) is odd.

408

Appendix A. Transforms, Identities, and Formulas

r Symmetry property for real signals

Suppose w(t) is real.
If w(t) = w(t), W (f ) is real.

(A.36)

If w(t) = w(t), W (f ) is purely imaginary.

r Time-shift property
With F {w(t)} = W (f ),
F {w(t t0 )} = W (f )ej2f t0 .

(A.37)

r Frequency scale property

With F {w(t)} = W (f ),
F {w(at)} =

1
W
a

 
f
.
a

(A.38)

r Dierentiation property
With F {w(t)} = W (f ),

dw(t)
= j2f W (f ).
dt
Convolution multiplication property
With F {wi (t)} = Wi (f ),

(A.39)

F {w1 (t) w2 (t)} = W1 (f )W2 (f )

(A.40)

F {w1 (t)w2 (t)} = W1 (f ) W2 (f ),

(A.41)

and

where the convolution operator is dened via


x()y( )d.
x() y()

(A.42)

r Parsevals theorem

With F {wi (t)} = Wi (f ),



w1 (t)w2 (t)dt =

r Final-value theorem

W1 (f )W2 (f )df.

(A.43)

With limt w(t) = 0 and w(t) bounded,

lim w(t) = lim j2f W (f ),

where F {w(t)} = W (f ).

f 0

(A.44)

Transforms, Identities, and Formulas

A.3

409

Energy and Power

r Energy of a continuous-time signal s(t) is


s2 (t)dt

E(s) =

(A.45)

if the integral is nite.

r Power of a continuous-time signal s(t) is

P (s) = lim

1
T

T /2

s2 (t)dt

(A.46)

T /2

if the limit exists.

r Energy of a discrete-time signal s[k] is

E(s) =

s2 [k]

(A.47)

if the sum is nite.

r Power of a discrete-time signal s[k] is

N
1  2
s [k]
N 2N

P (s) = lim

(A.48)

k=N

if the limit exists.

r Power spectral density

With input and output transforms X(f ) and Y (f ) of a linear lter with
impulse response transform H(f ) (such that Y (f ) = H(f )X(f )),
Py (f ) = Px (f )|H(f )|2 ,

(A.49)

where the power spectral density (PSD) is dened as

|XT (f )|2
T
T

Px (f ) = lim

(Watts/Hz),

(A.50)

where F {xT (t)} = XT (f ) and

 
t
,
xT (t) = x(t)
T

(A.51)

where () is the rectangular pulse (A.20).

A.4

Z-Transforms and Properties

r Denition of the Z-transform
X(z) = Z{x[k]} =


k=

x[k]z k

(A.52)

410

Appendix A. Transforms, Identities, and Formulas

r Time-shift property
With Z{x[k]} = X(z),
Z{x[k ]} = z X(z).

(A.53)

Z{ax1 [k] + bx2 [k]} = aX1 (z) + bX2 (z).

(A.54)

r Linearity property
With Z{xi [k]} = Xi (z),

r Final-value theorem for z-transforms

If X(z) converges for |z| > 1 and all poles of (z 1)X(z) are inside the unit
circle, then
lim x[k] = lim (z 1)X(z).

A.5

(A.55)

z1

Integral and Derivative Formulas

r Sifting property of impulse


r Schwarzs inequality





w(t)(t t0 )dt = w(t0 )

 2 

a(x)b(x)dx


|a(x)| dx

(A.56)

|b(x)| dx
2

(A.57)

and equality occurs only when a(x) = kb (x), where superscript indicates
complex conjugation (i.e., (a + jb) = a jb).
Leibnizs rule
$
%
b(x)
d
f (, x)d
a(x)
db(x)
da(x)
= f (b(x), x)
f (a(x), x)
dx
dx
dx
 b(x)
f (, x)
d
(A.58)
+
x
a(x)

r Chain rule of dierentiation

dw dy
dw
=
dx
dy dx

(A.59)

d
dy
dw
(wy) = w
+y
dx
dx
dx

(A.60)

r Derivative of a product

Transforms, Identities, and Formulas

411

r Derivative of signal raised to a power

dy
d n
(y ) = ny n1
dx
dx

(A.61)

d
dy
(cos(y)) = (sin(y))
dx
dx

(A.62)

dy
d
(sin(y)) = (cos(y))
dx
dx

(A.63)

(AT )T = A

(A.64)

(AB)T = B T AT

(A.65)

r Derivative of cosine

r Derivative of sine

A.6

Matrix Algebra
r Transpose transposed
r Transpose of a product
r Transpose and inverse commutativity
If A1 exists,

T

= A1 .

(A.66)

A1 A = AA1 = I.

(A.67)

AT

r Inverse identity
If A1 exists,

Simulating Noise

Noise generally refers to unwanted or undesirable signals that disturb or interfere

with the operation of a system. There are many sources of noise. In electrical
systems, there may be coupling with the power lines, lightning, bursts of solar
radiation, or thermal noise. Noise in a transmission system may arise from atmospheric disturbances, from other broadcasts that are not well shielded, and from
unreliable clock pulses or inexact frequencies used to modulate signals.
Whatever the physical source, there are two very dierent kinds of noise: narrowband and broadband. Narrowband noise consists of a thin slice of frequencies.
With luck, these frequencies will not overlap the frequencies that are crucial to
the communication system. When they do not overlap, it is possible to build lters that reject the noise and pass only the signal, analogous to the lter designed
in Section 7.2.3 to remove certain frequencies from the gong waveform. When
running simulations or examining the behavior of a system in the presence of
narrowband noise, it is common to model the narrowband noise as a sum of
sinusoids.
Broadband noise contains signicant amounts of energy over a large range of
frequencies. This is problematic because there is no obvious way to separate the
parts of the noise that lie in the same frequency regions as the signals from the
signals themselves. Often, stochastic or probabilistic models are used to characterize the behavior of systems under uncertainty. The simpler approach employed
here is to model the noise in terms of its spectral content. Typically, the noise v
will also be assumed to be uncorrelated with the signal w, in the sense that Rwv
of (8.3) is zero. The remainder of this section explores mathematical models of
(and computer implementations for simulations of) several kinds of noise that
are common in communications systems.
The simplest type of broadband noise contains all frequencies in equal
amounts. By analogy with white light, which contains all frequencies of visible light, this is called white noise. Most random-number generators, by default,
give (approximately) white noise. For example, the following Matlab code uses

413

Simulating Noise

Amplitude

10

20

30
40
50
(a) Time in seconds

Figure B.1 (a) The samples of a

random signal and (b) its (white)
spectrum.

60

Magnitude

5
0
1000
800
600
400
200
0

400

200
0
200
(b) Frequency in Hz

400

the function randn to create a vector with N normally distributed (or Gaussian)
random numbers.
Listing B.1. randspec.m spectrum of random numbers
N=216;
Ts = 0 . 0 0 1 ; t=Ts * ( 1 :N ) ;
x=randn ( 1 ,N ) ;
p l o t s p e c ( x , Ts ) ;

%
%
%
%

how many random numbers

d e f i n e a time v e c t o r
N random numbers
p l o t n o i s e and i t s spectrum

Running randspec.m gives a plot much like that shown in Figure B.1, though
details may change because the random numbers are dierent each time the
program is run.
The random numbers themselves fall mainly between 4, though most are less
than 2. The average (or mean) value is
m=

N
1 
x[k]
N

(B.1)

k=1

and is very close to zero, as can be veried by calculating

m=sum( x )/ length ( x ) .

The variance (the width, or spread of the random numbers) is dened by

v=

N
1 
(x[k] m)2
N

(B.2)

k=1

and can easily be calculated with the Matlab code

v=sum( ( xm) . * ( xm) ) / length ( x ) .

For randn, this is very close to 1.0. When the mean is zero, this variance is the
same as the power. Hence, if m=0, v=pow(x) also gives the variance.
The spectrum of a numerically generated white-noise sequence typically
appears as in the bottom plot of Figure B.1. Observe the symmetry in the spectrum (which occurs because the random numbers are real-valued). In principle,
the spectrum is at (all frequencies are represented equally), but in reality, any

414

Appendix B. Simulating Noise

given time the program is run, some frequencies appear slightly larger than others. In Figure B.1, there is such a spurious peak near 275 Hz, and a couple more
near 440 Hz. Verify that these spurious peaks occur at dierent frequencies each
time the program is run.
Exercise B.1. Use randspec.m to investigate the spectrum when dierent numbers of random values are chosen. Try N = 10, 100, 210, 218 . For each of the values
N, locate any spurious peaks. When the same program is run again, do they occur
at the same frequencies?1
Exercise B.2. Matlabs randn function is designed so that the mean is always
(approximately) zero and the variance is (approximately) unity. Consider a signal
dened by w = a randn + b; that is, the output of randn is scaled and oset.
What are the mean and variance of w? Hint: use (B.1) and (B.2). What values
must a and b have to create a signal that has mean 1.0 and variance 5.0?
Exercise B.3. Another Matlab function to generate random numbers is rand,
which creates numbers between 0 and 1. Try the code x=rand(1,N)-0.5 in
randspec.m, where the 0.5 causes x to have zero mean. What are the mean
and the variance of x? What does the spectrum of rand look like? Is it also
white? What happens if the 0.5 is removed? Explain what you see.
Exercise B.4. Create two dierent white signals w[k] and v[k] that are at least
N = 216 elements long.
a. For each j between 100 and +100, nd the cross-correlation Rwv [j] between
w[k] and v[k].
b. Find the autocorrelations Rw [j] and Rv [j]. What value(s) of j give the largest
autocorrelation?
Though many types of noise may have a wide bandwidth, few are truly white.
A common way to generate random sequences with (more or less) any desired
spectrum is to pass white noise through a linear lter with a specied passband.
The output then has a spectrum that coincides with the passband of the lter.
For example, the following program creates such colored noise by passing white
noise through a bandpass lter that attenuates all frequencies but those between
100 and 200 Hz:
Listing B.2. randcolor.m generating a colored noise spectrum
N=216;
Ts = 0 . 0 0 1 ; nyq =0.5/ Ts ;
s s f =(N/ 2 :N/2 1)/( Ts *N ) ;
1

% how many random numbers

% s a m p l i n g and n y q u i s t r a t e s
% frequency vector

Matlab allows control over whether the random numbers are the same each time using
the seed option in the calls to the random-number generator. Details can be found in the
help les for rand and randn.

Magnitude

Simulating Noise

800
600
400
200
0

Figure B.2 A white input signal

(top) is passed through a
bandpass lter, creating a
noisy signal with bandwidth
between 100 and 200 Hz.

(a) Spectrum of input random numbers

Magnitude

415

600
400
200
0

400
200
0
200
400
(b) Spectrum of output random numbers (Frequency in Hz)

x=randn ( 1 ,N ) ;
f b e =[0 100 110 190 200 nyq ] / nyq ;
damps=[0 0 1 1 0 0 ] ;
f l =70;
b=f i r p m ( f l , f b e , damps ) ;
y=f i l t e r ( b , 1 , x ) ;

%
%
%
%
%
%

N random numbers
define desired f i l t e r
des i r ed amplitudes
f i l t er size
design impulse response
f i l t e r x with b

Plots from a typical run of randcolor.m are shown in Figure B.2, which illustrates the spectrum of the white input and the spectrum of the colored output.
Clearly, the bandwidth of the output noise is (roughly) between 100 and 200 Hz.
Exercise B.5. Create a noisy signal that has no energy below 100 Hz. It should
then have (linearly) increasing energy from 100 Hz to the Nyquist rate at 500 Hz.
a. Design an appropriate lter using firpm. Verify its frequency response by
using freqz.
b. Generate a white noise and pass it through the lter. Plot the spectrum of
the input and the spectrum of the output.
Exercise B.6. Create two noisy signals w[k] and v[k] that are N = 216 elements
long. The bandwidths of both w[k] and v[k] should lie between 100 and 200 Hz
as in randcolor.m.
a. For each j between 100 and +100, nd the cross-correlation Rwv [j] between
w[k] and v[k].
b. Find the autocorrelations Rw [j] and Rv [j]. What value(s) of j give the largest
autocorrelation?
c. Are there any similarities between the two autocorrelations?
d. Are there any similarities between these autocorrelations and the impulse
response b of the bandpass lter?

Envelope of a Bandpass Signal

The envelope of a signal is a curve that smoothly encloses the signal, as shown in
Figure C.1. An envelope detector is a circuit (or computer program) that outputs
the envelope when the signal is applied at its input.
In early analog radios, envelope detectors were used to help recover the message from the modulated carrier, as discussed in Section 5.1. One simple design
includes a diode, capacitor, and resistor arranged as in Figure C.2. The oscillating signal arrives from an antenna. When the voltage is positive, current passes
through the diode, and charges the capacitor. When the voltage is negative,
the diode blocks the current, and the capacitor discharges through the resistor.
The time constants are chosen so that the charging of the capacitor is quick (so
that the output follows the upward motion of the signal), but the discharging is
relatively slow (so that the output decays slowly from its peak value). Typical
output of such a circuit is shown by the jagged line in Figure C.1, a reasonable
approximation to the actual envelope.
It is easy to approximate the action of an envelope detector. The essence
of the method is to apply a static nonlinearity (analogous to the diode in the
circuit) followed by a lowpass lter (the capacitor and resistor). For example, the
Matlab code in AMlarge.m on page 83 extracted the envelope using an absolutevalue nonlinearity and an LPF, and this method is also used in envsig.m.
Listing C.1. envsig.m envelope of a bandpass signal
time = . 3 3 ; Ts =1/10000;
t =0:Ts : time ; l e n t=length ( t ) ;
f c =1000; c=cos (2 * pi * f c * t ) ;
fm=10;
w=cos (2 * pi * fm * t ) . * exp(5 * t ) + 0 . 5 ;
x=c . * w ;
f b e =[0 0 . 0 5 0 . 1 1 ] ; damps=[1 1 0 0 ] ;
f l =100; b=f i r p m ( f l , f b e , damps ) ;
envx=(pi / 2 ) * f i l t e r ( b , 1 , abs ( x ) ) ;

%
%
%
%
%
%
%
%
%

sample i n t e r v a l & time

d e f i n e time v e c t o r
s i g n a l i s f a s t wave
f r e q f o r s l o w wave
t i m e s s l o w decay
with o f f s e t
lowpass f i l t e r design
i m p u l s e r e s p o n s e o f LPF
r e c t i f y f o r envelope

Suppose that a pure sine wave is input into this envelope detector. Then the
output of the LPF would be the average of the absolute value of the sine wave
(the integral of the absolute value of a sine wave over a period is /2). The factor

Envelope of a Bandpass Signal

417

Approximation of
envelope using
analog circuit

Upper
envelope

Figure C.1 The envelope

of a signal outlines the
extremes in a smooth
manner.

Signal
Lower
envelope

Figure C.2 A circuit that

Signal

Envelope

extracts the envelope from

a signal.

/2 in the denition of envx accounts for this factor so that the output rides
on crests of the wave. The output of envsig.m is shown in Figure C.3(a), where
the envelope signal envx follows the outline of the narrow-bandwidth passband
signal x, though with a slight delay. This delay is caused by the linear lter, and
can be removed by shifting the envelope curve by the group delay of the lter.
This is fl/2, half the length of the lowpass lter when designed using the firpm
command.
A more formal denition of envelope uses the notion of in-phase and quadrature components of signals to reexpress the original bandpass signal x(t) as the
product of a complex sinusoid and a slowly varying envelope function,
x(t) = Re{g(t)ej2fc t }.

(C.1)

The function g(t) is called the complex envelope of x(t), and fc is the carrier
frequency in Hz.
To see that this is always possible, consider Figure C.4. The input x(t) is
assumed to be a narrowband signal centered near fc (with support between
fc B and fc + B for some small B). Multiplication by the two sine waves
modulates this to a pair of signals centered at baseband and at 2fc . The LPF
removes all but the baseband, and so the spectra of both xc (t) and xs (t) are
contained between B to B. Modulation by the nal two sinusoids returns the
baseband signals to a region around fc , and adding them together gives exactly
the signal x(t). Thus, Figure C.4 represents an identity. It is useful because it
allows any passband signal to be expressed in terms of two baseband signals,
which are called the in-phase and quadrature components of the signal.

Amplitude

Appendix C. Envelope of a Bandpass Signal

Amplitude

418

1.5
1
0.5
0
0.5
1
1.5

1.5
1
0.5
0
0.5
1
1.5

Figure C.3 The envelope

0.05

0.1

0.05

0.1

0.15
0.2
(a)
Time

0.25

0.3

0.15
0.2
Time

0.25

0.3

(b)

smoothly outlines the

contour of the signal:
(a) shows the output of
envsig.m, while (b) shifts
the output to account for
the delay caused by the
linear lter.

Figure C.4 The envelope can

2 cos(2f0kT)

cos(2f0kT)
LPF

xc(t)
+

x(t)
LPF
2 sin(2f0kT)

xs(t)

sin(2f0kT)

x(t)

be written in terms of the

two baseband signals xc (t)
(the in-phase component)
and xs (t) (the quadrature
component). Assuming that
the lowpass lters are
perfect, this represents an
identity; x(t) at the input
equals x(t) at the output.

Symbolically, the signal x(t) can be written

x(t) = xc (t) cos(2fc t) xs (t) sin(2fc t),
where
xc (t) = LPF{2x(t) cos(2fc t)},

(C.2)

xs (t) = LPF{2x(t) sin(2fc t)}.

(C.3)

Applying Eulers identity (A.1) then shows that the envelope g(t) can be
expressed in terms of the in-phase and quadrature components as
*
g(t) = x2c (t) + x2s (t).
Any physical (real-valued) bandlimited waveform can be represented as in
(C.1) and so it is possible to represent many of the standard modulation schemes
in a unied notation.

Envelope of a Bandpass Signal

419

For example, consider the case in which the complex envelope is a scaled
version of the message waveform (i.e., g(t) = Ac w(t)). Then
x(t) = Re{Ac w(t)ej2fc t }.
Using ejx = cos(x) j sin(x),
x(t) = Re{w(t)[Ac cos(2fc t) + jAc sin(2fc t)]}
= w(t)Ac cos(2fc t),
which is the same as AM with a suppressed carrier from Section 5.2.
AM with a large carrier can also be written in the form of (C.1) with g(t) =
Ac [1 + w(t)]. Then
x(t) = Re{Ac [1 + w(t)]ej2fc t }
= Re{Ac ej2fc t + Ac w(t)ej2fc t }
= Ac cos(2fc t) + w(t)Ac cos(2fc t).
The envelope g(t) is real in both of these cases when w(t) is real.
When the envelope g(t) = x(t) + jy(t) is complex-valued, x(t) in (C.1)
becomes
x(t) = Re{(x(t) + jy(t))ej2fc t }.
With ejx = cos(x) + j sin(x),
x(t) = Re{x(t) cos(2fc t) + jx(t) sin(2fc t) + jy(t) cos(2fc t) + j 2 y(t) sin(2fc t)}
= x(t) cos(2fc t) y(t) sin(2fc t).
This is the same as the quadrature modulation of Section 5.3.
Exercise C.1. Replace the filter command with the filtfilt command and
rerun envsig.m. Observe the eect of the delay. Read the Matlab help le for
filtfilt, and try to adjust the programs so that the outputs coincide. Hint:
you will need to change the lter parameters as well as the decay of the output.
Exercise C.2. Replace the absolute-value nonlinearity with a rectifying nonlinearity

x(t) t 0
x(t) =
0
t < 0,
which more closely simulates the action of a diode. Mimic the code in envsig.m
to create an envelope detector. What is the appropriate constant that must be
used to make the output smoothly touch the peaks of the signal?
Exercise C.3. Use envsig.m and the following code to nd the envelope of a
signal. Can you see how to write these three lines of code in one (complex-valued)
line?

420

Appendix C. Envelope of a Bandpass Signal

xc=f i l t e r ( b , 1 , 2 * x . * cos (2 * pi * f c * t ) ) ;
x s=f i l t e r ( b , 1 , 2 * x . * s i n (2 * pi * f c * t ) ) ;
envx=abs ( xc+sqrt ( 1) * x s ) ;

% in phase component
% q u a d r a t u r e component
% envelope of s i g n a l

Exercise C.4. For those who have access to the Matlab signal-processing toolbox, an even simpler syntax for the complex envelope is
envx=abs(hilbert(x));
Can you gure out why the Hilbert transform is useful for calculating the
envelope?
Exercise C.5. Find a signal x(t) for which all the methods of envelope detection
fail to provide a convincing envelope. Hint: try signals that are not narrowband.

Relating the Fourier Transform to

the DFT

Most people are quite familiar with time-domain thinking: a plot of voltage
versus time, how stock prices vary as the days pass, a function that grows (or
shrinks) over time. One of the most useful tools in the arsenal of an electrical
engineer is the idea of transforming a problem into the frequency domain. Sometimes this transformation works wonders; what at rst seemed intractable is now
obvious at a glance. Much of this appendix is about the process of making the
transformation from time into frequency, and back again from frequency into
time. The primary mathematical tool is the Fourier transform (and its discretetime counterparts).

D.1

The Fourier Transform and Its Inverse

By denition, the Fourier transform of a time function w(t) is

W (f ) =
w(t)ej2f t dt,

(D.1)

which appeared earlier in Equation (2.1). The inverse Fourier transform is


W (f )ej2f t df.
(D.2)
w(t) =

Observe that the transform is an integral over all time, while the inverse transform is an integral over all frequency; the transform converts a signal from time
into frequency, while the inverse converts from frequency into time. Because the
transform is invertible, it does not create or destroy information. Everything
about the time signal w(t) is contained in the frequency signal W (f ) and vice
versa.
The integrals (D.1) and (D.2) do not always exist; they may fail to converge
or they may become innite if the signal is bizarre enough. Mathematicians
have catalogued exact conditions under which the transforms exist, and it is a

422

Appendix D. Relating the Fourier Transform to the DFT

reasonable engineering assumption that any signal encountered in practice fullls

these conditions.
Perhaps the most useful property of the Fourier transform (and its inverse) is
its linearity. Suppose that w(t) and v(t) have Fourier transforms W (f ) and V (f ),
respectively. Then superposition suggests that the function s(t) = aw(t) + bv(t)
should have transform S(f ) = aW (f ) + bV (f ), where a and b are any complex
numbers. To see that this is indeed the case, observe that


s(t)ej2f t dt =
(aw(t) + bv(t))ej2f t dt
S(f ) =



w(t)ej2f t dt + b
v(t)ej2f t dt
=a

= aW (f ) + bV (f ).
What does the transform mean? Unfortunately, this is not immediately apparent from the denition. One common interpretation is to think of W (f ) as
describing how to build the time signal w(t) out of sine waves (more accurately,
out of complex exponentials). Conversely, w(t) can be thought of as the unique
time waveform that has the frequency content specied by W (f ).
Even though the time signal is usually a real-valued function, the transform
W (f ) is, in general, complex-valued due to the complex exponential ej2f t
appearing in the denition. Thus W (f ) is a complex number for each frequency
f . The magnitude spectrum is a plot of the magnitude of the complex numbers
W (f ) as a function of f , and the phase spectrum is a plot of the angle of the
complex numbers W (f ) as a function of f .

D.2

The DFT and the Fourier Transform

This section derives the DFT as a limiting approximation of the Fourier transform, showing the relationship between the continuous-time and discrete-time
transforms.
The Fourier transform cannot be applied directly to a waveform that is dened
only on a nite interval [0, T ]. But any nite-length signal can be extended to
innite length by assuming it is zero outside of [0, T ]. Accordingly, consider the
windowed waveform

#


w(t) 0 t T
t (T /2)
= w(t)
,
ww (t) =
T
0
t otherwise
where is the rectangular pulse (2.8). The Fourier transform of this windowed
(nite-support) waveform is

 T
ww (t)ej2f t dt =
w(t)ej2f t dt.
Ww (f ) =
t=

t=0

423

Relating the Fourier Transform to the DFT

Approximating the integral at f = n/T and replacing the dierential dt with t

(= T /N ) allows it to be approximated by the sum

 T
N
1


w(t)e2jf t dt

w(kt)ej2(n/T )k(T /N ) t
0

f =n/T

k=0

= t

N
1


w(kt)ej(2/N )nk ,

k=0

where the substitution t kt is used. Identifying w(kt) with w[k] gives

Ww (f )|f =n/T t W [n].
As before, T is the time window of the data record, and N is the number of
data points. t (= T /N ) is the time between samples (or the time resolution),
which is chosen to satisfy the Nyquist rate so that no aliasing will occur. T is
selected for a desired frequency resolution f = 1/T ; that is, T must be chosen
large enough so that f is small enough. For a frequency resolution of 1 Hz, a
second of data is needed. For a frequency resolution of 1 kHz, 1 ms of data is
needed.
Suppose N is to be selected so as to achieve a time resolution t = 1/(f ),
where > 2 causes no aliasing (i.e., the signal is bandlimited to f ). Suppose
T is specied to achieve a frequency resolution 1/T that is times the signals
highest frequency, so T = 1/(f ). Then the (required) number of data points
N , which equals the ratio of the time window T to the time resolution t, is
/.
For example, consider a waveform that is zero for all time before Td /2, when
it becomes a sine wave lasting until time Td /2. This switched sinusoid can be
modeled as
 
 
t
t
A sin(2f0 t) =
A cos(2fo t /2).
w(t) =
Td
Td
From (2.9), the transform of the pulse is
  
sin(f Td )
t
.
F
= Td
Td
f Td
Using the frequency-translation property, the transform of the switched sinusoid
is
 

1
sin((f f0 )Td )
sin((f + f0 )Td )
j/2
j/2
e
Td
+e
Td
W (f ) = A
,
2
(f f0 )Td
(f + f0 )Td
which can be simplied (using ej/2 = j and ej/2 = j) to


jATd sin((f + f0 )Td ) sin((f f0 )Td )
.

W (f ) =
2
(f + f0 )Td
(f f0 )Td

(D.3)

This transform can be approximated numerically, as in the following program

switchsin.m. Assume the total time window of the data record of N = 1024

424

Appendix D. Relating the Fourier Transform to the DFT

1
60
0.5
40

0
0.5
1
4

20
2
0
2
4
Time/seconds
(a) Switched sinusoid

0.5

0.5

0.4

0.4

0.3

0.3

0.2

0.2

0.1

0.1

500
1000 1500
Bin number
(b) Raw DFT magnitude

Figure D.1 Spectrum of the

switched sinusoid
calculated using the DFT:
(a) the time waveform,
(b) the raw magnitude
data, (c) the magnitude
spectrum, and (d) zoomed
into the magnitude
spectrum.

0
0
0
50 100 40 20
0
20
40
100 50
Frequency/Hz
Frequency/Hz
(c) Magnitude spectrum (d) Zoom into magnitude spectrum

samples is T = 8 s and that the underlying sinusoid of frequency f0 = 10 Hz is

switched on for only the rst Td = 1 s.
Listing D.1. switchsin.m spectrum of a switched sine
Td=1;
% p u l s e width [ Td/ 2 ,Td / 2 ]
N=1024;
% number o f data p o i n t s
f =10;
% frequency of sine
T=8;
% t o t a l time window
t r e z=T/N; f r e z =1/T ;
% time and f r e q r e s o l u t i o n
w=zeros ( 1 ,N ) ;
% v e c t o r f o r f u l l data r e c o r d
w(N/2Td/ ( t r e z * 2 ) + 1 :N/2+Td/ ( 2 * t r e z ))= s i n ( t r e z * ( 1 : Td/ t r e z ) * 2 * pi * f ) ;
dftmag=abs ( f f t (w ) ) ;
% magnitude o f spectrum o f w
s p e c=t r e z * [ dftmag ( (N/ 2 )+ 1 :N) , dftmag ( 1 :N / 2 ) ] ;
s s f=f r e z * [ (N/ 2 ) + 1 : 1 : (N / 2 ) ] ;
% frequency vector
plot ( t r e z * [ length (w) / 2 + 1 : length (w) / 2 ] , w, - ) ; % p l o t ( a )
plot ( dftmag , - ) ;
% pl ot ( b)
plot ( s s f , s p e c , - ) ;
% plot ( c)

Plots of the key variables are shown in Figure D.1. The switched sinusoid w
is shown plotted against time, and the raw spectrum dftmag is plotted as
a function of its index. The proper magnitude spectrum spec is plotted as a
function of frequency, and the nal plot shows a zoom into the low-frequency
region. In this case, the time resolution is t = T /N = 0.0078 s and the frequency
resolution is f = 1/T = 0.125 Hz. The largest allowable f0 without aliasing is
N/(2T ) = 64 Hz.
Exercise D.1. Rerun the preceding program with T=16, Td=2, and f=5. Comment on the location and the width of the two spectral lines. Can you nd
particular values so that the peaks are extremely narrow? Can you relate the
locations of these narrow peaks to (D.3)?

Power Spectral Density

One way of classifying and measuring signals and systems is by their power (or
energy), and the amount of power (or energy) in various frequency regions. This
section denes the power spectral density, and shows how it can be used to
measure the power in signals, to measure the correlation within a signal, and to
talk about the gain of a linear system. In Software Receiver Design, power
spectral density is used mainly in Chapter 11 in the discussion of the design of
matched ltering.
The (time) energy of a signal was dened in (A.45) as the integral of the signal
squared, and Parsevals theorem (A.43) guarantees that this is the same as the
total energy measured in frequency

|W (f )|2 df,
E=

where W (f ) = F {w(t)} is the Fourier transform of w(t).

When E is nite, w(t) is called an energy waveform. But E is innite for many
common signals in communications such as the sine wave and the sinc functions.
In this case, the power, as dened in (A.46), is

1 T /2
|w(t)|2 dt,
(E.1)
P = lim
T T t=T /2
which is the average of the energy, and which can be used to measure the signal.
Signals for which P is nonzero and nite are called power waveforms.
Dene the truncated waveform
 
t
wT (t) = w(t)
,
T
where () is the rectangular pulse (2.8) that is 1 between T /2 and T /2, and
is zero elsewhere. When w(t) is real-valued, (E.1) can be rewritten

1
P = lim
wT2 (t)dt.
T T t=

426

Appendix E. Power Spectral Density

Parsevals theorem (A.43) shows that this is the same as


 

1
|WT (f )|2
|WT (f )|2 df =
P = lim
lim
df,
T T f =
T
f = T
where WT (f ) = F {wT (t)}. The power spectral density (PSD) is then dened as
|WT (f )|2
T
T

Pw (f ) = lim

(Watts/Hz),

which allows the total power to be written


P =
Pw (f )df.

(E.2)

f =

Observe that the PSD is always real and nonnegative. When w(t) is real-valued,
then the PSD is symmetric, Pw (f ) = Pw (f ).
The PSD can be used to reexpress the autocorrelation function (the correlation
of w(t) with itself),

1 T /2
Rw ( ) = lim
w(t)w(t + )dt,
T T T /2
in the frequency domain. This is the continuous-time counterpart of the crosscorrelation (8.3) with w = v. First, replace with . Now the integrand is a
convolution, and so the Fourier transform is the product of the spectra. Hence,


1
w(t) w(t)
F {Rw ( )} = F {Rw ( )} = F lim
T T
1
1
= lim
F {w(t) w(t)} = lim
|W (f )|2 = Pw (f ).
T T
T T
Thus, the Fourier transform of the autocorrelation function of w(t) is equal to
the power spectral density of w(t),1 and it follows that

P =
Pw (f )df = Rw (0),

which says that the total power is equal to the autocorrelation at = 0.

The PSD can also be used to quantify the power gain of a linear system.
Recall that the output y(t) of a linear system is given by the convolution of the
impulse response h(t) with the input x(t). Since convolution in time is the same
as multiplication in frequency, Y (f ) = X(f )H(f ). Assuming that H(f ) has nite
energy, the PSD of y is
Py (f ) = lim

1
1
|YT (f )|2 = lim
|XT (f )|2 |H(f )|2 = Px (f )|H(f )|2 ,
T T
T

(E.3)

This
( is known as the WienerKhintchine theorem, and it formally requires that
Rw ( )d be nite; that is, the correlation between w(t) and w(t + ) must die away

as gets large.

Power Spectral Density

427

where yT (t) = y(t)(t/T ) and xT (t) = x(t)(t/T ) are truncated versions of y(t)
and x(t). Thus, the PSD of the output is precisely the PSD of the input times
the magnitude of the frequency response (squared), and the power gain of the
linear system is exactly |H(f )|2 for each frequency f .

The Z -Transform

This appendix presents background material that is useful when designing equalizers and when analyzing feedback systems such as the loop structures that occur
in many adaptive elements. The Z and the s play more than an orthographic
role since the Z-transform (the discrete-time counterpart of the s-domain Laplace
transform) can be used to represent a linear system such as a channel model or
a lter in a concise way. The frequency response of these models can easily be
derived using a simple graphical technique that also provides insight into the
inverse model. This can be useful in visualizing equalizer design as in Chapter
13, and the same techniques are useful in the analysis of systems such as the
PLLs of Chapter 10. The open-eye criterion provides a way of determining
how good the design is.

F.1

Z-Transforms
Fundamental to any digital signal is the idea of the unit delay, a time delay T of
exactly one sample interval. There are several ways to represent this mathematically, and this section uses the Z-transform, which is closely related to a discrete
version of the Fourier transform. Dene the variable z to represent a (forward)
time shift of one sample interval. Thus, zu(kT ) u((k + 1)T ). The inverse is the
backward time shift z 1 u(kT ) u((k 1)T ). These are most commonly written
without explicit reference to the sampling rate as
zu[k] = u[k + 1] and z 1 u[k] = u[k 1].
For example, the FIR lter with input u[k] and output
y[k] = u[k] + 0.6u[k 1] 0.91u[k 2]
can be rewritten in terms of the time-delay operator z as
y[k] = (1 + 0.6z 1 0.91z 2)u[k].

The Z -Transform

429

Similarly, the IIR lter

y[k] = a1 y[k 1] + a2 y[k 2] + b0 u[k] + b1 u[k 1]
can be rewritten
(1 a1 z 1 a2 z 2 )y[k] = (b0 + b1 z 1 )u[k].
As will become clear, it is often quite straightforward to implement and analyze
a system described using polynomials in z.
Formally, the Z-transform is dened much like the transforms of Chapter 5.
The Z-transform of a sequence y[k] is

Y (z) = Z{y[k]} =

y[k]z k .

(F.1)

k=

Though it might not at rst be apparent, this denition corresponds to the

intuitive idea of a unit delay. The Z-transform of a delayed sequence is
Z{y[k ]} =

y[k ]z k .

k=

Applying the change of variable k = j (so that k = j + ), this can be

rewritten



y[j]z (j+) = z
y[j]z j = z Y (z). (F.2)
Z{y[k ]} =
j+=

j+=

In words, the Z-transform of the time-shifted sequence y[k ] is z times the

Z-transform of the original sequence y[k]. Observe the similarity between this
property and the time-delay property of Fourier transforms, Equation (A.37).
This similarity is no coincidence; formally substituting z ej2f and t0
turns (F.2) into (A.37).
In fact, most of the properties of the Fourier transform and the DFT have a
counterpart in Z-transforms. For instance, it is easy to show from the denition
(F.1) that the Z-transform is linear; that is,
Z{ay[k] + bu[k]} = aY (z) + bU (z).
Similarly, the product of two Z-transforms is given by the convolution of the
time sequences (analogous to (7.2)), and the ratio of the Z-transform of the
output to the Z-transform of the input is a (discrete-time) transfer function.
For instance, the simple two-tap nite-impulse-response dierence equation,
y[k] = u[k] bu[k 1],

(F.3)

430

Appendix F: The Z -Transform

can be represented in transfer-function form by taking the Z-transform of both

sides of the equation, applying (F.2), and using linearity. Thus,
Y (z) Z{y[k]} = Z{u[k] bu[k 1]} = Z{u[k]} Z{bu[k 1]}
= U (z) bz 1 U (z) = (1 bz 1 )U (z),
which can be solved algebraically for
H(z) =

Y (z)
zb
= 1 bz 1 =
.
U (z)
z

(F.4)

H(z) is called the transfer function of the lter (F.3).

There are two types of singularities that a z-domain transfer function may
have. Poles are those values of z that make the magnitude of the transfer function
innite. The transfer function in (F.4) has a pole at z = 0. Zeros are those values
of z that make the magnitude of the frequency response equal to zero. The
transfer function in (F.4) has one zero at z = b. There are always exactly as
many poles as there are zeros in a transfer function, though some may occur at
innite values of z. For example, the transfer function (z a)/1 has one nitevalued zero at z = a and a pole at z = .
A z-domain discrete-time system transfer function is called minimum phase
(maximum phase) if it is causal and all of its singularities are located inside
(outside) the unit circle. If some singularities are inside and others outside the
unit circle, the transfer function is mixed phase. If it is causal and all of the poles
of the transfer function are strictly inside the unit circle (i.e., if all the poles have
magnitudes less than unity), then the system is stable, and a bounded input
always leads to a bounded output. For example, the FIR dierence equation
y[k] = u[k] + 0.6u[k 1] 0.91u[k 2]
has the transfer function
Y (z)
= 1 + 0.6z 1 0.91z 2
H(z) =
U (z)
(z 0.7)(z + 1.3)
.
z2
This is mixed phase and stable, with zeros at z = 0.7 and 1.3 and two poles at
z = 0.
It can also be useful to represent IIR lters analogously. For example, Section
7.2.2 shows that the general form of an IIR lter is
= (1 0.7z 1 )(1 + 1.3z 1 ) =

y[k] = a0 y[k 1] + a1 y[k 2] + + an y[k n 1]

+ b0 x[k] + b1 x[k 1] + + bm x[k m].
This can be rewritten more concisely as
(1 a0 z 1 a1 z 2 an z n )Y (z) = (1 + b0 z 1 + b1 z 2 + + bm z m )U (z)

The Z -Transform

431

and rearranged to give the transfer function

1 + b0 z 1 + b1 z 2 + + bm z m
Y (z)
=
U (z)
1 a0 z 1 a1 z 2 an z n
z m + b0 z m1 + b1 z m2 + + bm
= n
.
z a0 z n1 a1 z n2 an

H(z) =

(F.5)

An interesting (and sometimes useful) feature of the Z-transform is the relationship between the asymptotic value of the sequence in time and a limit of the
transfer function.
Final-Value Theorem for Z-transforms: If X(z) converges for |z| > 1 and
all poles of (z 1)X(z) are inside the unit circle, then
lim x[k] = lim (z 1)X(z).

z1

(F.6)

Exercise F.1. Use the denition of the Z-transform to show that the transform
is linear; that is, show that Z{ax[k] + bu[k]} = aX(z) + bU (z).
Exercise F.2. Find the z-domain transfer function of the system dened by
y[k] = b1 u[k] + b2 u[k 1].
a.
b.
c.
d.
e.
f.

What are the poles of the

What are the zeros?
Find limk y[k].
For what values of b1 and
For what values of b1 and
For what values of b1 and

transfer function?

b2 is the system stable?

b2 is the system minimum phase?
b2 is the system maximum phase?

Exercise F.3. Find the z-domain transfer function of the system dened by
y[k] = ay[k 1] + bu[k 1].
a.
b.
c.
d.
e.
f.

What are the poles of the transfer function?

What are the zeros?
Find limk y[k].
For what values of a and b is the system stable?
For what values of a and b is the system minimum phase?
For what values of a and b is the system maximum phase?

Exercise F.4. Find the z-domain transfer function of the lter dened by y[k] =
y[k 1] 2u[k] + (2 )u[k 1].
a.
b.
c.
d.

What are the poles of the transfer function?

What are the zeros?
Find limk y[k].
Is this system stable?

432

Appendix F: The Z -Transform

F.2

Sketching the Frequency Response from the Z-Transform

A complex number = a + jb can be drawn in the complex plane as a vector
from the origin to the point (a, b). Figure F.1 is a graphical illustration of the
dierence between two complex numbers , which is equal to the vector
drawn from to . The magnitude is the length of this vector, and the angle
is measured counterclockwise from the horizontal drawn to the right of to the
direction of , as shown.
As with Fourier transforms, the discrete-time transfer function in the z-domain
can be used to describe the gain and phase that a sinusoidal input of frequency
(in radians/second) will experience when passing through the system. With
transfer function H(z), the frequency response can be calculated by evaluating
the magnitude of the complex number H(z) at all points on the unit circle, that
is, at all z = ejT . (T has units of seconds/sample.)
For example, consider the transfer function H(z) = z a. At z = ej0T = 1
(zero frequency), H(z) = 1 a. As the frequency increases (as increases), the
test point z = ejT moves along the unit circle (think of this as the in
Figure F.1). The value of the frequency response at the test point H(ejT ) is
the dierence between this and the zero of H(z) at z = a (which corresponds
to in Figure F.1).
Suppose that 0 < a < 1. Then the distance from the test point to the zero is
smallest when z = 1, and increases continuously as the test point moves around
the circle, reaching a maximum at T = radians. Thus, the frequency response
is highpass. The phase goes from 0 to 180 as T goes from 0 to . On the other
hand, if 1 < a < 0, then the system is lowpass.
More generally, consider any polynomial transfer function
aN z N + aN 1 z N 1 + + a2 z 2 + a1 z + a0 .
Im

| + |

Figure F.1 Graphical calculation of

Re()Re()
Im()Im()

Re

the dierence between two complex

numbers.

The Z -Transform

180o = radians
Nyquist sampling
rate

Frequency of
test points given
in radians

z2
b1

Zeros

b3

b2

a3

a1
a2

z1

DC or zero
frequency

433

Figure F.2 Suppose a transfer

function has three zeros. At any

frequency test point (specied
in radians around the unit
circle), the magnitude of the
transfer function is the product
of the distances from the test
point to the zeros.

This can be factored into a product of N (possibly complex) roots:

H(z) = gN
i=1 (z i ).

(F.7)

Accordingly, the magnitude of this FIR transfer function at any value z is the
product of the magnitudes of the distances from z to the zeros. For any test
point on the unit circle, the magnitude is equal to the product of all the distances
from the test point to the zeros. An example is shown in Figure F.2, where a
transfer function has three zeros. Two test points are shown at frequencies
corresponding to (approximately) 15 and 80 . The magnitude at the rst test
point is equal to the product of the lengths a1 , a2 , and a3 , while the magnitude at
the second is b1 b2 b3 . Qualitatively, the frequency response begins at some value
and slowly decreases in magnitude until it nears the second test point. After this,
it rises. Accordingly, this transfer function is a notch lter.
Exercise F.5. Consider the transfer function (z a)(z b) with 1 > a > 0 and
0 > b > 1. Sketch the magnitude of the frequency response, and show that it
has a bandpass shape over the range of frequencies between 0 and radians.
As another example, consider a ring of equally spaced zeros in the complex-z
plane. The resulting frequency-response magnitude will be relatively at because,
no matter where the test point is taken on the unit circle, the distance to the
zeros in the ring of zeros is roughly the same. As the number of zeros in the ring
decreases (increases), scallops in the frequency-response magnitude will become
more (less) pronounced. This is true irrespective of whether the ring of transferfunction zeros is inside or outside the unit circle. Of course, the phase curves will
be dierent in the two cases.
Exercise F.6. Sketch the frequency response of H(z) = z a when a = 2.
Sketch the frequency response of H(z) = z a when a = 2.
Exercise F.7. Sketch the frequency responses of

434

Appendix F: The Z -Transform

a.
b.
c.
d.

H(z) = (z 1)(z 0.5)

H(z) = z 2 2z + 1
H(z) = (z 2 2z + 1)(z + 1)
H(z) = g(z n 1) for g = 0.1, 1.0, 10 and n = 2, 5, 25, 100.

Exercise F.8. A second-order FIR lter has two zeros, both with positive real
parts. TRUE or FALSE: The lter is a highpass lter.
Exercise F.9. TRUE or FALSE: The linear, time-invariant lter with impulse
response
f [k] =

[k j]

j=

is a highpass lter.
Of course, these frequency responses can also be evaluated numerically. For
instance, the impulse response of the system described by H(z) = 1 + 0.6z 1
0.91z 2 is the vector h=[1 0.6 -0.91]. Using the command freqz(h) draws
the frequency response.
Exercise F.10. Draw the frequency response for each of the systems H(z) in
Problem F.7 using Matlab.
Exercise F.11. Develop a formula analogous to (F.7) that holds when H(z)
contains both poles and zeroes as in (F.5).
a. Sketch the frequency response for the system of Exercise F.3 when a = 0.5
and b = 1.
b. Sketch the frequency response for the system of Exercise F.3 when a = 0.98
and b = 1.
c. Sketch the frequency response for the system of Exercise F.3 when a = 1 and
b = 1.
If the transfer function includes nite-valued poles, the gain of the transfer
function is divided by the product of the distances from a test point on the unit
circle to the poles. The counterclockwise angle from the positive horizontal at
each pole location to the vector pointing from there to the test point on the unit
circle is subtracted in the overall phase formula. The point of this technique is
not to carry out complex calculations better left to computers, but to learn to
reason qualitatively using plots of the singularities of transfer functions.

The Z -Transform

F.3

435

Measuring Intersymbol Interference

The ideas of frequency response and dierence equations can be used to interpret and analyze properties of the transmission system. When all aspects of the
system operate well, quantizing the received signal to the nearest element of the
symbol alphabet recovers the transmitted symbol. This requires (among other
things) that there is no signicant multipath interference. This section uses the
graphical tool of the eye diagram to give a measure of the severity of the intersymbol interference. In Section 11.3, the eye diagram was introduced as a way
to visualize the intersymbol interference caused by various pulse shapes. Here,
the eye diagram is used to help visualize the eects of intersymbol interference
caused by multipath channels such as (13.2).
For example, consider a binary 1 source s[k] and a three-tap FIR channel
model that produces the received signal
r[k] = b0 s[k] + b1 s[k 1] + b2 s[k 2].
This is shown in Figure F.3, where the received signal is quantized using the
sign operator in order to produce the binary sequence y[k], which provides an
estimate of the source. Depending on the values of the bi , this estimate may, but
might not, accurately reect the source.
Suppose b1 = 1 and b0 = b2 = 0. Then r[k] = s[k 1] and the output of the
decision device is, as desired, a replica of a delayed version of the source
(i.e., y[k] = sign{s[k 1]} = s[k 1]). Like the eye diagrams of Chapter 9, which
are open whenever the intersymbol interference admits perfect reconstruction
of the source message, the eye is said to be open.
If b0 = 0.5, b1 = 1, and b2 = 0, the system is r[k] = 0.5s[k] + s[k 1] and there
are four possibilities: (s[k], s[k 1]) = (1, 1), (1, 1), (1, 1), or (1, 1), for
which r[k] is 1.5, 0.5, 0.5, and 1.5, respectively. In each case, sign{r[k]} =
s[k 1]. The eye is still open.
Now consider b0 = 0.4, b1 = 1, and b2 = 0.2. The eight possibilities for
(s[k], s[k 1], s[k 2]) in
r[k] = 0.4s[k] + s[k 1] 0.2s[k 2]
are (1, 1, 1), (1, 1, 1), (1, 1, 1), (1, 1, 1), (1, 1, 1), (1, 1, 1), (1, 1, 1),
and (1, 1, 1). The resulting choices for r[k] are 1.2, 1.6, 0.8, 0.4, 0.4,
0.8, 1.6, and 1.2, respectively, with the corresponding s[k 1] of 1, 1, 1,
1, 1, 1, 1, and 1. For all of the possibilities, y[k] = sign{r[k]} = s[k 1].
Received
signal

Channel
Source
s

b0 + b1

z1 +

b2

z2

Decision
device
sign(.)

Reconstructed
source
estimate
y

Figure F.3 Channel and

binary decision device.

436

Appendix F: The Z -Transform

Furthermore, y[k] = s[k] and y[k] = s[k 2] across the same set of choices. The
eye is still open.
Now consider b0 = 0.5, b1 = 1, and b2 = 0.6. The resulting r[k] are 0.9, 2.1,
1.1, 0.1, 0.1, 1.1, 2.1, and 0.9, with s[k 1] = 1, 1, 1, 1, 1, 1, 1, and
1. Out of these eight possibilities, two cause sign{r[k]} = s[k 1]. (Neither s[k]
nor s[k 2] does better.) The eye is closed.
This can be explored in Matlab using the program openclosed.m, which
denes the channel in b and implements it using the filter command. After
passing through the channel, the binary source becomes multivalued, taking on
values b1 b2 b3 . Typical outputs of openclosed.m are shown in Figure F.4
for channels b=[0.4 1 -0.2] and [0.5 1 -0.6]. In the rst case, four of the
possible values are above zero (when b2 is positive) and four are below (when
b2 is negative). In the second case, there is no universal correspondence between
the sign of the input data and the sign of the received data y. This is the purpose
of the nal for statement, which counts the number of errors that occur at each
delay. In the rst case, there is a delay that causes no errors at all. In the second
case, there are always errors.
Listing F.1. openclosed.m draw eye diagrams
b=[0.4 1 0.2];
% def i ne channel
m=1000; s=sign ( randn ( 1 ,m) ) ;
% binary input of length m
r=f i l t e r ( b , 1 , s ) ;
% output o f channel
y=sign ( r ) ;
% quantization
f o r sh =0:5
% e r r o r at d i f f e r e n t d el ay s
e r r ( sh +1)=0.5 *sum( abs ( y ( sh +1:m) s ( 1 :msh ) ) ) ;
end

In general for the binary case, if for some i


|bi | >
|bj |,
j=i

then such incorrect decisions cannot occur. The greatest distortion occurs at the
boundary between the open and closed eyes. Let be the index at which the
impulse response has its largest coecient (in magnitude), so |b | |bi | for all
i. Dene the open-eye measure for a binary 1 input

i= |bi |
.
OEM = 1
|b |
For b0 = 0.4, b1 = 1, and b2 = 0.2, OEM = 1 (0.6/1) = 0.4. This value is
how far from zero (i.e., crossing over to the other source-alphabet value) the
equalizer output is in the worst case (as can be seen in Figure F.4). Thus, errorfree behavior could be assured as long as all other sources of error are smaller
than this OEM. For the channel [0.5, 1, 0.6], the OEM is negative, and the
eye is closed.

The Z -Transform

437

Eye is open for the channel [0.4, 1, 0.2]

Amplitude

2
1
0
1
2

100 200 300 400 500 600 700 800 900 1000

Amplitude

Eye is closed for the channel [0.5, 1, 0.6]

3
2
1
0
1
2
3

Figure F.4 Eye diagrams

for two channels.

100 200 300 400 500 600 700 800 900 1000
Symbol number

If the source is not binary, but instead takes on maximum (smax ) and minimum
(smin ) magnitudes, then, as a worst-case measure,

( i= |bi |)smax
OEM = 1
.
|b |smin
As dened, OEM is always less than unity, with this value achieved only in the
trivial case that all |bi | are zero for i = and |b | > 0. Thus,

r OEM > 0 is good (i.e., open eye)

r OEM < 0 is bad (i.e., closed eye)
The OEM provides a way of measuring the interference from a multipath
channel. It does not measure the severity of other kinds of interference such as
noise or in-band interferences caused by (say) other users.
Exercise F.12. Use openclosed.m to investigate the channels
a. b = [0.3, 0.3, 0.3, 1, 0.1]
b. b = [0.1, 0.1, 0.1, 0.1, 0.1, 0.1, 0.1]
c. b = [1, 2, 3, 10, 3, 2, 1]
For each channel, is the eye open? If so, what is the delay associated with the
open eye? What is the OEM measure in each case?
Exercise F.13. Modify openclosed.m so that the received signal is corrupted
by a (bounded uniform) additive noise with maximum amplitude s. How does
the equivalent of Figure F.4 change? For what values of s do the channels in
Problem F.12 have an open eye? For what values of s does the channel b = [0.1,

438

Appendix F: The Z -Transform

0.1, 10, 0.1] have an open eye? Hint: use 2*s*(rand-0.5) to generate the
noise.
Exercise F.14. Modify openclosed.m so that the input uses the source alphabet
1, 3. Are any of the channels in Problem F.12 open eye? Is the channel b=
[0.1, 0.1, 10, 0.1] open eye? What is the OEM measure in each case?
When a channel has an open eye, all the intersymbol interference can be
removed by the quantizer. But when the eye is closed, something more must
be done. Opening a closed eye is the job of an equalizer, and is discussed at
length in Chapter 13.
Exercise F.15. The sampled impulse response from symbol values to the output of the downsampler in the receiver depends on the baud timing of the
downsampler. For example, let h(t) be the impulse response from symbol values
m(kT ) to downsampler output y(kT ). The sampled impulse response sequence
is h(t)|t=kT + for integers k and symbol period T , where is the selected baudtiming oset. Consider the nonzero impulse response values for two dierent
choices for ,
h(t)|t=kT +1 = {0.04, 1.00, 0.52, 0.32, 0.18},
h(t)|t=kT +2 = {0.01, 0.94, 0.50, 0.26, 0.06}
for k = 0, 1, 2, 3, 4.
a. With = 1 and a binary source alphabet 1, is the system from the source
symbols to the downsampler output open-eye? Clearly justify your answer.
b. With = 2 and a binary source alphabet 1, is the system from the source
symbols to the downsampler output open-eye? Clearly justify your answer.
c. Consider ltering the downsampler output with an equalizer with transfer
function F (z) = (z )/z to produce soft decisions y[k]. Thus, the discretetime impulse response from the symbols m[k] to the soft decisions y[k] is the
convolution of the sampled impulse response from the symbol values to the
downsampler output with the equalizer impulse response. With = 1 and a
four-element source alphabet {1, 3} can the system from the source symbols to the equalizer output be made open-eye by selecting = 0.6? Explain.

F.4

Analysis of Loop Structures

The Z-transform can also be used to understand the behavior of linearized loop
structures such as the PLL and the dual-loop variations. This section begins by
studying the stability and tracking properties of the system in Figure 10.22 on
page 224, which is the linearization of the general structure in Figure 10.20. The
nal-value theorem (F.6) shows that, for certain choices of the transfer function

The Z -Transform

439

G(z) = B(z)/A(z), the loop can track both step inputs (which correspond to
phase osets in the carrier) and ramp inputs (which correspond to frequency
osets in the carrier). This provides analytical justication of the assertions in
Chapter 10 on the tracking capabilities of the loops.
As discussed in Section 10.6.3, Figure 10.22 is the linearization of Figure 10.20.
The input to the linearized system is 2 and the goal is to choose the LPF G(z) so
that 2 converges to 2. This can be rewritten as the desire to drive e = 2 2
to zero. Observe that a phase oset in the carrier corresponds to a nonzero and
a step input applied to the linearized system. A frequency oset in the carrier
corresponds to a ramp input (i.e., a line with slope proportional to the frequency
oset) applied to the linearized system.
Accordingly, the behavior of the system can be studied by nding the transfer
function E(z)/(z) from to e. The asymptotic value of E(z)/(z) can be
found using the nal-value theorem with (z) equal to a step and with (z)
equal to a ramp. If these are zero, the linearized system converges to the input,
i.e., e converges to zero and so converges to . This demonstrates the desired
tracking behavior.
is /(z 1) and
The transfer function from the output of the block G(z) to [k]
is G(z)/(z 1). With G(z) = B(z)/A(z)
so the transfer function from e to [k]
expressed as a ratio of two polynomials in z, the transfer function from to e is
E(z)
(z 1)A(z)
=
.
(z)
(z 1)A(z) + 2B(z)

(F.8)

To apply the nal-value theorem, let [k] be a step with height ,


0 k<0
[k] =
k 0.
Then the Z-transform of is (z) = z/(z 1) and
E(z) =

zA(z)
(z 1)A(z)
(z) =
.
(z 1)A(z) + 2B(z)
(z 1)A(z) + 2B(z)

Applying the nal-value theorem shows that

z1
(z 1)A(z)
zA(z)
= lim
z1
z (z 1)A(z) + 2B(z) z1 (z 1)A(z) + 2B(z)

lim e[k] = lim

as long as B(1) = 0 and the roots of

(z) = (z 1)A(z) + 2B(z)

(F.9)

are strictly inside the unit circle. Since the numerator is zero, limk e[k] = 0.
In other words, as time k increases, the estimated phase converges to [k]. Thus
the system can track a phase oset in the carrier.
When [k] is a ramp with slope ,

0
k<0
[k] =
k k 0,

440

Appendix F: The Z -Transform

and so (z) = z/(z 1)2 . In this case,

E(z) =

zA(z)
(z 1)((z 1)A(z) + 2B(z))

and the nal-value theorem shows that

lim e[k] = lim

z1

A(1)
z1
A(z)
E(z) = lim
=
z1 (z 1)A(z) + 2B(z)
z
2B(1)

as long as B(1) = 0 and the roots of (F.9) are strictly inside the unit circle.
If A(1) = 0, limk e[k] is nonzero and decreases with larger . If A(1) = 0,
converges to the ramp [k]. This means that the
limk e[k] is zero and [k]
loop can also asymptotically track frequency osets in the carrier.
Changing shifts the closed-loop poles in (F.9). This impacts both the stabil locks
ity of the roots of (z) and the decay rate of the transient response as [k]
onto and tracks [k]. Poles closer to the origin of the z-plane correspond to faster
decay of the transient repsonse. Poles closer to the origin of the z-plane than to
z = 1 tend to de-emphasize the lowpass nature of the G(z), thereby weakening the suppression of any high-frequency disturbances from the input. Thus,
the location of the closed-loop poles compromises between improved removal of
broadband noises and a faster decay of the transient error.
The choice of G(z) which includes a pole at z = 1 (i.e., an integrator) results
in a PLL with type-II tracking capability, a designation commonly used in
feedback control systems where polynomial tracking is a frequent task. The name
arises from the ability to track a type-I polynomial (i = 1 in input ki with k
the time index) with zero error asymptotically and a type-II polynomial (k 2 )
with asymptotically constant (and nite) oset.
Exercise F.16. Suppose that a loop such as Figure 10.20 has G(z) = b/(z a).
a.
b.
c.
d.

What is the corresponding (z) of (F.9)?

Does this loop track phase osets? Does it track frequency osets?
For what value of does (z) become unstable?
Find a value of for which the roots of (z) are complex-valued. What does
this imply about the convergence of the loop?
e. Describe the behavior of the loop when a = 1.

Exercise F.17. Suppose that a loop such as Figure 10.20 has G(z) = b(z
c)/(z 1).
a.
b.
c.
d.

What is the corresponding (z) of (F.9)?

Does this loop track phase osets? Does it track frequency osets?
For what value of does (z) become unstable?
Find a value of for which the roots of (z) are complex-valued. What does
this imply about the convergence of the loop?

The Z -Transform

441

Exercise F.18. Suppose that a loop such as Figure 10.20 has G(z) = b(z
1)2 /(z 1 ).
a. What is the corresponding (z) of (F.9)?
b. Does this loop track phase osets? Does it track frequency osets?
c. For what value of does (z) become unstable?
Exercise F.19. The general loop in Figure 10.20 is simulated in pllgeneral.m.
a. Implement the lter G(z) from Exercise F.16 and verify that it behaves as
predicted.
b. Implement the lter G(z) from Exercise F.17 and verify that it behaves as
predicted.
c. Implement the lter G(z) from Exercise F.19 and verify that it behaves as
predicted.

Averages and Averaging

There are two results in this appendix. The rst section argues that averages
(whether implemented as a simple sum, as a moving average, or in recursive form)
have an essentially lowpass character. This is used repeatedly in Chapters
6, 10, 12, and 13 to study the behavior of adaptive elements by simplifying
the objective function to remove extraneous high-frequency signals. The second
result is that the derivative of an average (or an LPF) is almost the same as
the average (or LPF) of the derivative. This approximation is formalized in
(G.12) and is used throughout Software Receiver Design to calculate the
derivatives that occur in adaptive elements such as the phase-locked loop, the
automatic gain control, output energy maximization for timing recovery, and
various equalization algorithms.

G.1

Averages and Filters

There are several kinds of averages. The simple average [N ] of a sequence of N
numbers [i] is
[N ] = avg{[i]} =

N
1 
[i].
N i=1

(G.1)

For instance, the average temperature last year can be calculated by adding up
the temperature on each day, and then dividing by the number of days.
When talking about averages over time, it is common to emphasize recent
data and to de-emphasize data from the distant past. This can be done using a
moving average of length P , which has a value at time k of
1
[k] = avg{[i]} =
P

k

i=kP +1

[i].

(G.2)

Averages and Averaging

443

This can also be implemented as a nite-impulse-response lter

1
1
1
[k] + [k 1] + + [k P + 1].
(G.3)
P
P
P
Instead of averaging the temperature over the whole year all at once, a moving
average over a month (P = 30) nds the average over each consecutive 30-day
period. This would show, for instance, that it is very cold in Wisconsin in the
winter and hot in the summer. The simple annual average, on the other hand,
would be more useful to the Wisconsin tourist bureau, since it would show a
moderate yearly average of about 50 degrees.
Closely related to these averages is the recursive summer,
[k] =

[i] = [i 1] + [i] for i = 1, 2, 3, . . . ,

(G.4)

which adds up each new element of the input sequence [i], scaled by . Indeed,
if the recursive lter (G.4) has = 1/N and is initialized with [0] = 0, then
[N ] is identical to the simple average in (G.1).
Writing these averages in the form of the lters (G.4) and (G.3) suggests the
following question: what kind of lters are these? The impulse response h[k] of
the moving average lter is

k<0

0
1
h[k] =
0k<P

P
0
k P,
which is essentially a rectangle shape in time. Accordingly, the frequency
response is sinc-shaped, from (A.20). Thus, the averaging lter passes very
low frequencies and attenuates high frequencies. It thus has a lowpass character,
though it is far from an ideal LPF.
The impulse response for the simple recursive lter (G.4) is

0
k<0
h[k] =
k 0.
This is also a rectangle that widens as k increases, which again represents a
lter with a lowpass character. This can be seen using the techniques of Appendix
F by observing that the transfer function of (G.4) has a single pole at 1 which
causes the magnitude of the frequency response to decrease as the frequency
increases. Thus, averages such as (G.1), moving-average lters such as (G.3),
and recursive lters such as (G.4) all have a lowpass character.

G.2

Derivatives and Filters

Averages and lowpass lters occur within the denitions of the performance functions associated with many adaptive elements. For instance, the AGC of Chapter
6, the phase-tracking algorithms of Chapter 10, the timing-recovery methods of

444

Appendix G. Averages and Averaging

Chapter 12, and the equalizers of Chapter 13 all involve LPFs, averages, or both.
Finding the correct form for the adaptive updates requires taking the derivative
of ltered and averaged signals. This section shows when it is possible to commute the two operations, that is, when the derivative of the ltered (averaged)
signal is the same as a ltering (averaging) of the derivative. The derivative is
taken with respect to some variable , and the key to the commutativity is how
enters the ltering operation. Sometimes the derivative is taken with respect
to time, sometimes it is taken with respect to a coecient of the lter, and
sometimes it appears as a parameter within the signal.
When the derivative is taken with respect to time, the LPF and/or average
commute with the derivative; that is,

d
d
=
LPF{}
(G.5)
LPF
d
d
and


avg

d
d

d
avg{},
d

(G.6)

where is the signal and represents time. This is a direct consequence of linearity; the LPF and the derivative are both linear operations. Since linear operations commute, so do the lters (averages) and the derivatives. This is demonstrated using the code in dlpf.m in which a random signal s is passed through
an arbitrary linear system dened by the impulse response h. The derivative is
approximated in dlpf.m using the diff function, and the calculation is done two
ways: rst by taking the derivative of the ltered signal, and then by ltering the
derivative. Observe that the two methods give the same output after the lters
have settled.
Listing G.1. dlpf.m dierentiation and ltering commute
s=randn ( 1 , 1 0 0 ) ;
h=randn ( 1 , 1 0 ) ;
d l p f s=d i f f ( f i l t e r ( h , 1 , s ) ) ;
l p f d s=f i l t e r ( h , 1 , d i f f ( s ) ) ;
d l p f s l p f d s

%
%
%
%
%

g e n e r a t e random data
an a r b i t r a r y i m p u l s e r e s p o n s e
take deriv of f i l t e r e d input
f i l t e r th e d e r i v o f i n p u t
compare th e two

When the derivative is taken with respect to a coecient (tap weight) of the
lter, then (G.5) does not hold. For example, consider the time-invariant linear
lter
[k] =

P
1


bi [k i],

i=0

which has impulse response [bP 1 , . . . , b1 , b0 ]. If the bi are chosen so that [k]
represents a lowpass ltering of the [k], then the notation
[k] = LPF{[k]}

Averages and Averaging

445

is appropriate, while if bi = 1/P , then [k] = avg{[k]} might be more apropos.

In either case, consider the derivative of [k] with respect to a parameter bj :
d
d[k]
=
(b0 [k] + b1 [k 1] + + bP 1 [k P + 1])
dbj
dbj
db1 [k] db2 [k 1]
dbP 1 [k P + 1]
=
+
+ +
.
dbj
dbj
dbj

(G.7)

Since all the terms bi [k i] are independent of bj for i = j, all of the derivatives
dbi [k i]/dbj are zero. For i = j,
dbj
dbj [k j]
= [k j]
= [k j],
dbj
dbj
so
d[k]
dLPF{[k]}
=
= [k j].
dbj
dbj

(G.8)

On the other hand, LPF {d[k]/dbj } = 0 because [k] is not a function of bj .

The derivative and the lter do not commute and (G.5) (with = bj ) does not
hold.
An interesting and useful case is when the signal that is to be ltered is a
function of . Let [, k] be the input to the lter that is explicitly parameterized
by both and by time k. Then
[k] = LPF{(, k)} =

P
1


bi (, k i).

i=0

The derivative of [k] with respect to is



P 1
P
1

dLPF{(, k)}
d(, k i)
d 
d(, k)
bi (, k i) =
bi
=
= LPF
.
d
d i=0
d
d
i=0
Thus, (G.5) holds in this case.

Example G.1. This example is reminiscent of the phase-tracking algorithms in

Chapter 10. Let = and (, k) = (, k) = sin(2f kT + ). Then
P 1
P
1

dLPF{(, k)}
d 
d
=
bi sin(2(k i)T + ) =
bi sin(2(k i)T + )
d
d i=0
d
i=0


P
1

d(, k)
=
bi cos(2(k i)T + ) = LPF
.
(G.9)
d
i=0

Example G.2. This example is reminiscent of the equalization algorithms that

appear in Chapter 13 in which the signal (, k) is formed by ltering a signal

446

Appendix G. Averages and Averaging

u[k] that is independent of . To be precise, let = a1 and (, k) = (a1 , k) =

a0 u[k] + a1 u[k 1] + a2 u[k 2]. Then dLPF{(a1 , k)}/da1 is given by
=

P 1
d 
bi (a0 u[k i] + a1 u[k i 1] + a2 u[k i 2])
da1 i=0
P
1


d
(a0 u[k i] + a1 u[k i 1] + a2 u[k i 2])
da1
i=0


P
1

d(a1 , k)
bi u[k i 1] = LPF
=
.
da1
=

bi

i=0

The transition between the second and third equality mimics the transition from
(G.7) to (G.8), with u playing the role of and a1 playing the role of bj .

G.3

Dierentiation Is a Technique, Approximation Is an Art

When (the variable with respect to which the derivative is taken) is not a
function of time, then the derivatives can be calculated without ambiguity or
approximation, as was done in the previous section. In most of the applications in
Software Receiver Design, however, the derivative is being calculated for the
express purpose of adapting the parameter; that is, with the intent of changing
so as to maximize or minimize some performance function. In this case, the
derivative is not straightforward to calculate, and it is often simpler to use an
approximation.
To see the complication, suppose that the signal is a function of time k and
the parameter and that is itself time-dependent. Then it is more proper to
use the notation ([k], k), and to take the derivative with respect to [k]. If it
were simply a matter of taking the derivative of ([k], k) with respect to [k],
then there would be no problem, since

d(, k) 
d([k], k)
=
.
(G.10)
d[k]
d =[k]
When taking the derivative of a ltered version of the signal ([k], k), however, all the terms are not exactly of this form. Suppose, for example, that
[k] = LPF{([k], k)} =

P
1


bi ([k i], k i)

i=0

is a ltering of and the derivative is to be taken with respect to [k]:

P
1

d[k]
d([k i], k i)
=
.
bi
d[k]
d[k]
i=0

Averages and Averaging

447

Only the rst term in the sum has the form of (G.10). All others are of the form
d([k i], k i)
,
d[k]
with i = 0. If there were no functional relationship between [k] and [k i],
this derivative would be zero, and

d(, k) 
d[k]
= b0
.
d[k]
d =[k]
But, of course, there generally is a functional relationship between at dierent
times, and proper evaluation of the derivative requires that this relationship be
taken into account.
The situation that is encountered repeatedly throughout Software Receiver
Design occurs when [k] is dened by a small-stepsize iteration; that is,
[k] = [k 1] + ([k 1], k 1),

(G.11)

where ([k 1], k 1) is some bounded signal (with bounded derivative) which
may itself be a function of time k 1 and the state [k 1]. A key feature of
(G.11) is that is a user-choosable stepsize parameter. As will be shown, when
a suciently small is chosen, the derivative d[k]/d[k] can be approximated
eciently as
P
1
P
1


d([k i], k i)
d([k i], k i)
dLPF{([k], k)}
=

bi
bi
d[k]
d[k]
d[k i]
i=0
i=0
#


d(, k) 
,
(G.12)
= LPF
d =[k]

which nicely recaptures the commutativity of the LPF and the derivative as in
(G.5). A special case of (G.12) is to replace LPF with avg.
The remainder of this section provides a detailed justication for this approximation and provides two detailed examples. Other examples appear throughout
Software Receiver Design.
As a rst step, consider the derivative which appears in the second term of
the sum in (G.12). This can be rewritten using the chain rule (A.59) as
d([k 1], k 1)
d([k 1], k 1) d[k 1]
=
.
d[k]
d[k 1]
d[k]
Rewriting (G.11) as [k 1] = [k] ([k 1], k 1) yields
d([k 1], k 1) d([k] ([k 1], k 1))
d[k 1]
d[k]




d(, k 1) 
d([k 1], k 1)
=
.
1


d
d[k]
=[k1]

(G.13)

448

Appendix G. Averages and Averaging

Applying similar logic to the derivative of shows that

d([k 1], k 1) d[k 1]
d([k 1], k 1)
=
d[k]
d[k 1]
d[k]
d([k 1], k 1) d([k] ([k 1], k 1))
=
d[k 1]
d[k]



d(, k 1) 
d([k 1], k 1)
.
=
1

d
d[k]
=[k1]
While this may appear at rst glance to be a circular argument (since
d([k 1], k 1)
d[k]
appears on both sides), it can be solved algebraically as

d(, k 1) 

d
0
d([k 1], k 1)
=[k1]


,
=
d(, k 1) 
d[k]
1 + 0
1+

d

(G.14)

=[k1]

where


d(, k 1) 
.
0 =

d
=[k1]

(G.15)

Substituting (G.14) back into (G.13) and simplifying yields


d([k 1], k 1)
1
d(, k 1) 
.
=

d[k]
d
1
+
0
=[k1]
The point of this calculation is that, since the value of is chosen by the user,
it can be made as small as needed to ensure that

d([k 1], k 1)
d(, k 1) 

.

d[k]
d
=[k1]

Following the same basic steps for the general delay term in (G.12) shows that

d(, k n) 
d([k n], k n)
(1 n ),
=

d[k]
d
=[kn]
where

n1

0
n =
j
1
1 + 0
j=1

is dened recursively with 0 given in (G.15) and 1 = 0 /(1 + 0 ). For small

, this implies that

d(, k n) 
d([k n], k n)

(G.16)

d[k]
d
=[kn]

Averages and Averaging

449

for each n. Combining these yields the approximation (G.12).

Example G.3. Example G.1 assumes that the phase angle is xed, even though
the purpose of the adaptation in a phase-tracking algorithm is to allow to follow
a changing phase. To investigate the time-varying situation, let [k] = [k] and
([k], k) = ([k], k) = sin(2f kT + [k]). Suppose also that the dynamics of
are given by
[k] = [k 1] + ([k 1], k 1).
Using the approximation (G.16) yields
P 1
d 
dLPF{([k], k)}
=
bi ([k i], k i)
d[k]
d[k] i=0

=
=

P 1
d 
bi sin(2(k i)T + [k i])
d[k]

P
1

i=0
P
1


i=0

bi

d
sin(2(k i)T + [k i])
d[k]

bi

d
sin(2(k i)T + [k i])
d[k i]

i=0
P
1


bi cos(2(k i)T + [k i])

#

d(, k) 
.
= LPF
d =[k]
i=0

(G.17)

Example G.4. Example G.2 assumes that the parameter a1 of the linear lter
is xed, even though the purpose of the adaptation is to allow a1 to change
in response to the behavior of the signal. Let ([k], k) be formed by ltering
a signal u[k] that is independent of [k]. To be precise, let [k] = a1 [k] and
([k], k) = (a1 [k], k) = a0 [k]u[k] + a1 [k]u[k 1] + a2 [k]u[k 2]. Suppose also
that the dynamics of a1 [k] are given by
a1 [k] = a1 [k 1] + (a1 [k 1], k 1).

450

Appendix G. Averages and Averaging

Then the approximation (G.16) yields

P 1
dLPF{(a1 [k], k)}
d 
bi (a1 [k i], k i)
=
da1 [k]
da1 [k] i=0

P 1
d 
bi (a0 [k i]u[k i] + + a2 [k i]u[k i 2])
da1 [k] i=0
P
1


bi

i=0
P
1


d
(a0 [k i]u[k i] + + a2 [k i]u[k i 2])
da1 [k]
P 1

 da1 [k i]u[k i 1]
da1 [k i]u[k i 1]

bi
da1 [k]
da1 [k i]
i=0
i=0

#

P
1

d(a1 , k) 
bi u[k i 1] = LPF
=
.
da1 

i=0

bi

a1 =a1 [k]

The B 3IG Transmitter

In Chapter 9, the Basic Black Box Impairment Generator of Figure 9.22 (on page
187) was described as a routine that transforms a Matlab script specifying the
operation of the transmitter into the (received) signal that appears at the input of
the receiver. This appendix opens up the black box, shining light on the internal
operation of the B 3IG.
The B 3IG is implemented in Matlab as the routine BigTransmitter.m, and
it allows straightforward modeling of any (or all) of the possible impairments
discussed throughout Software Receiver Design, including carrier-frequency
oset, baud-timing osets, and frequency-selective and time-varying channels,
as well as channel noise. Since many of the impairments and nonidealities that
arise in a communication system occur in the channel and RF front end, B 3IG
is more than a transmitter: it includes the communication channel and receiver
RF front end as well. An overview of the B 3IG is shown in Figure H.1.
The B 3IG architecture expands on the simplied communication system of
Chapter 9 and has more options than the M6 transmitter of Chapter 15. Some
of the additional features are as follows.
Support for multiple users. The transmitter generates a signal that may contain
information intended for more than one receiver.
Message for
User 1
Message for
Transmitter
User 2

...

Channel Between
Base Station
and User n

Sampled IF
Receiver
for User n

Received
Digital
Signal

Message for
User K
Figure H.1 The B 3IG picture: an overview of the transmission system which can be

used with multiple users, and has numerous ways to model impairments of the
received signal.

452

Appendix H. The B 3IG Transmitter

Figure H.2 A base station transmitting to several

mobile users within a single cell. Users within
the cell may use a time-delay protocol whereby
each user is assigned a sequential time slot, while
adjacent cells operate using nearby frequencies.

Assumption that the receiver employs a sampled intermediate-frequency architecture. Section 6.2 presented several choices for performing downsampling in
a receiver. In modern communication systems, where carrier frequencies can
be on the order of several gigahertz, it is rarely feasible to build a receiver
front end that samples at the Nyquist rate. The B 3IG transmitter generates a
received signal that has been brought down to a lower, intermediate frequency
(IF) and then sampled at some specied rate assumed to be lower than the
Nyquist rate of the IF.
Insertion of periodic preamble data. In practice, transmission of some known
preamble data is required so that the receiver can learn how to combat the
various impairments and nonidealities. In addition, the preamble permits the
receiver to identify the start of a frame or transmission.
The B 3IG is intended to mimic the base station in a cellular system in which
the geographical coverage area is split into hexagonally shaped cells as in Figure H.2. Thus there are many users in the same geographical region all wishing
to transmit and receive simultaneously. How is a given receiver able to separate its message from the composite received signal? The key is to separate the
transmission either in time or in frequency (or using a mixture of time and
frequency called the code space). These separation techniques are called timedivision multiple access (TDMA), frequency-division multiple access (FDMA),
and code-division multiple access (CDMA).
Within a single cell, the transmitter implements multiple users that send messages by sequencing them in time. In TDMA, where the signal intended for each
user is transmitted during a designated time slot, all users employ the same band
of frequencies. The frame structure for the B 3IG TDMA is shown in Figure H.3.
This may be contrasted with the FDMA situation where users each employ their
own frequency range, as was discussed in Figure 2.3 on page 22.
Typically, adjacent cells transmit on dierent frequencies, though frequencies
may be reused in cells that are suciently far away. Adjacent cells may use
frequencies that are close to each otherand may even overlap slightly due to
relaxed spectral masks. Consequently, it is not unreasonable to expect some

The B 3IG Transmitter

4(Ns + KNp) Symbols per Frame

... User K Preamble User 1 User 2 ... User K Preamble User 1 ...
Np Characters
in Preamble

Ns Characters
per User Slot

453

Figure H.3 The

time-division multiple
access (TDMA) frame
structure used in the B 3IG.
The indicated sizes are
valid for the 8-bit ASCII
encoding.

interference from adjacent cells. The B 3IG makes it easy to study the eects of
this kind of interference.
In code-division multiple access (CDMA), each user is assigned a code that is
mutually orthogonal to all of the other users. All users transmit at the same time
and in the same frequency range, and the decoding relies on the orthogonality of
the codes. The B 3IG does not currently implement a CDMA protocol. Adding a
CDMA component to the receiver would make a great course project!

H.1

Constructing the Received Signal

The B 3IG transmitter (i.e., the base station) simulates K simultaneous users
using TDMA. Each receiver (i.e., each mobile device in Figure H.2) retains and
decodes only its intended portion of the signal. For single-user operation, K = 1.
Each of the three blocks in Figure H.1 is itself built from several components
that are familiar from previous chapters. The system that generates the analog received signal (i.e., the actual transmitter and channel) is shown in block
diagram form in Figure H.4. This is, of course, the marginally magical M6 transmitter from Chapter 15. When the signal arrives at the front end of the receiver,
it is transformed into a sampled IF signal as in Figure H.5.
The original message in Figure H.4 is a character string of English text. Each
character is mapped into a sequence of 4-PAM symbols. The mappings may be
made using the Matlab function letters2pam.m, which converts each letter
into an 8-bit binary string, and hence to four 4-PAM symbols as in Example
8.1. Alternatively, the mappings may be made with the more terse text2bin.m,

Bits
Text
message Characterto-binary
conversion

Symbols
s[i]
Scaling
TDMA user
multiplexing

Trigger

Baseband
signal
s(t)
Pulse
shape
P( f )

Transmitted
passband
signal

Modulation
including
phase noise

Channel

Adjacent Broadband
users
noise
Analog
received
signal

1
iTt + t

Figure H.4 The signal-ow diagram of the B 3IG transmitter closely follows that of the

M6 transmitter of Figure 15.1 on page 343.

454

Appendix H. The B 3IG Transmitter

mixer
Preselect
Filter

r[k]
IF Filter

AGC

kTs

oscillator

Figure H.5 The front end

of the B 3IG transmission
system lters for the
desired frequency range,
adjusts the gain, and
modulates to the
intermediate frequency.
After a nal ltering, the
samples give the received
signal r.

which converts each letter into a 7-bit binary string, thus assigning seven 4-PAM
symbols to every pair of letters. This mapping is familiar from Example 8.2.
In order to decode the message at the receiver, the recovered symbols must be
properly grouped, the start of each group must be located, and the desired user
data must be extracted. To aid this TDMA frame synchronization, a preamble
is inserted in the symbol stream at the start of every frame. The preamble is
a short text message that starts each frame and it is assumed to be known at
the receiver. Like the message, it is encoded in 4-PAM symbols. The unknown
messages for each user are placed in sequence after the preamble, as in Figure
H.3.
As indicated in Figure H.4, pulses are initiated at intervals of Tt seconds,
and each is scaled by the 4-PAM symbol value. This translates the discrete-time
symbol sequence s[i] (composed of the messages interleaved with the preamble)
into a continuous time signal

s[i](t iTt t ).
s(t) =
i

The actual transmitter symbol period Tt is required to be within some tolerance

of the nominal symbol period T , but the B 3IG transmitter allows the introduction of a nonideal clock that is slightly dierent from the specied (true) clock
at the receiver. The transmitter-symbol-period clock is assumed to be steady
enough that the timing oset t and its period Tt are eectively time-invariant
over the duration of a single frame.
The pulse-shaping lter P (f ) is a square-root raised cosine lter symmetrically
truncated to a specied number of symbol periods. The rollo factor of the
pulse-shaping lter is xed within some range and is known at the receiver,
though it could take on dierent values with dierent transmissions. The pulseshaped signal is then converted to the radio frequency and passed through the
channel. Since the receiver is assumed to employ a sampled IF architecture, the
IF completely species the behavior of the signal in the frequency domain. While
it might seem preferable to generate simulated signals at the radio frequency (to
more accurately model the physical system), the number of samples needed is so
large that computer memory would be quickly exhausted.

The B 3IG Transmitter

455

The channel may be near ideal, i.e., a unit gain multi-symbol delay, or it may
have signicant intersymbol interference. In addition it can be time-varying, and
the B 3IG code allows two ways to introduce time variation. The rst permits
specication of the channel impulse response c1 at the start of the transmission
and the channel impulse response c2 at the end of the transmission. At any
time in between, the actual impulse response moves linearly from c1 to c2 . The
second method of introducing time variation is to model the channel variation
as a random walk. The variance of the random-walk process can be specied.
In either case, the impulse response of the channel is unknown at the receiver,
though an upper bound on its delay spread may be available in practice. There
are also other disturbances that may be present, including interference from
adjacent channels and narrowband interferers. These disturbances can also be
controlled using the B 3IG transmitter, and they are assumed to be unknown at
the receiver.
The achieved intermediate frequency is required to be within some userspecied tolerance of its assigned value. The carrier phase (t) is unknown to the
receiver and may vary over time, albeit slowly, due to phase noise. This means
that the phase of the intermediate-frequency signal presented to the receiver
sampler may also vary.
The front end of the sampled IF receiver, as shown in Figure H.5, consists
of a low-noise amplier, a preselect lter, an AGC, a mixer for downconversion
to IF, and the sampler. The preselect lter partially attenuates adjacent FDM
user bands. The automatic gain control is presumed locked and xed over each
transmission, and outputs a signal whose average power is unity. The free-running
sampler frequency fs needs to be well above twice the baseband bandwidth of
the user of interest to allow proper functioning of the baseband analog signal
interpolator in the DSP timer in the receiver. However, fs need not be twice
the highest frequency of the IF signal. This implies that the sampled received
signal has replicated the spectrum of the user transmission at the output of the
front-end analog downconverter lowpass lter to frequencies between zero and
IF.

H.2

Matlab Code for the Notorious B 3IG

H.2.1

Function Call
To generate the received signal, the B 3IG transmitter is called by
[r, s]=BigTransmitter(m, frameParams, rfParams, chanParams);
where m is the message (or messages, in the case of multiple users) to be
transmitted, frameParams is a structure containing the parameters relating to
the frame structure, rfParams contains the analog/RF-related parameters, and

456

Appendix H. The B 3IG Transmitter

chanParams contains the channel parameters. There are two output arguments:
r is the received sampled signal and s is the 4-PAM symbols which were transmitted. Thus, the length of s is approximately equal to the length of r times the
eective oversampling factor.

H.2.2

The Input Parameters

In order to control all these features of the B 3IG, various parameters are needed.
An overview of these parameters is given in Figure H.6, and more details are as
follows.
m: A character-string matrix, with dimensions equal to the number of users by
the message length, containing the ASCII messages to be transmitted. If there
is only a single user, this is a single string (row vector).

Description Symbol Matlab variable

message
m
number of characters per user per frame
Ns
frameParams.userDataLength
character string of preamble text
frameParams.preamble
enable or disable channel coding
frameParams.chanCodingFlag
7-bit or 8-bit ASCII representation
frameParams.bitEncodingFlag
channel coecients at start of transmission
c1
chanParams.c1
c2
channel coecients at end of transmission
chanParams.c2
variance of time-varying channel
chanParams.randomWalkVariance
SNR SNR chanParams.SNR
power of adjacent interferer 1
chanParams.adjacentUser1Power
carrier frequency of adjacent interferer 1
chanParams.adjacentUser1f if
channel of adjacent interferer 1
chanParams.adjacentUser1Chan
power of adjacent interferer 2
chanParams.adjacentUser2Power
carrier frequency of adjacent interferer 2
chanParams.adjacentUser2f if
channel of adjacent interferer 2
chanParams.adjacentUser2Chan
frequency of narrowband interferer
chanParams.NBIfreq
power of narrowband interferer
chanParams.NBIPower
sampler frequency
fs
rfParams.f s
nominal symbol period
Tt
rfParams.T t
% error of transmitter symbol period
rfParams.T t err
assigned intermediate frequency
f IF
rfParams.f if
% error transmitter intermediate frequency
rfParams.f if err
variance of phase noise
rfParams.phaseNoiseVariance
SRRC lter length
rfParams.SRRCLength
SRRC pulse-shape rollo factor

rfParams.SRRCrolloff
Figure H.6 B 3IG parameters passed via the stuctures frameParams, rfParams, and

chanParams.

The B 3IG Transmitter

457

frameParams.userDataLength: The number of text characters per user per

frame which (together with frameParams.preamble) eectively determines
the length of a frame. See Figure H.3.
frameParams.preamble: A character string (i.e., a row vector) containing the
preamble text which is presumed known at all receivers.
frameParams.chanCodingFlag: This parameter can be set to either 0 or 1. If set
to 0, the transmitter sends uncoded bits. If set to 1, the transmitter encodes
the data using the binary block (5,2) channel code described in Section 14.6.1.
frameParams.bitEncodingFlag: This parameter can be set to either 0 or 1,
and controls how the transmitter translates text into bits. If set to 0, the
transmitter uses an 8-bit representation of the ASCII code, and employs
letters2pam.m. If set to 1, the transmitter uses the 7-bit representation of the
ASCII code, and employs text2bin.m. See Examples 8.1 and 8.2 for details.
chanParams.c1: The impulse response of the time-varying channel at the start
of the transmission is specied as a row vector.
chanParams.c2: The impulse response of the time-varying channel at the end of
the transmission is also a row vector. The actual channel varies linearly over
the course of the transmission from the impulse response in chanParams.c1
to that in chanParams.c2. Hence, if chanParams.c1 = chanParams.c2 and
chanParams.randomWalkVariance = 0, the channel will be time invariant.
chanParams.randomWalkVariance: In addition to being able to control the evolution of the channel using chanParams.c1 and chanParams.c2, this parameter species a random variation in the channel taps using a random walk.
The parameter controls the variance of the random-walk process.
chanParams.SNR: The signal-to-noise ratio of the signal. Note that it controls
only the amount of AWGN present in the band of the transmitted signal.
chanParams.adjacentUser1Power: The relative power of the interference caused
by the rst adjacent user, specied in dB. Thus, when set to 0 dB the interferer
will have the same power as the transmitter. When set to -Inf, the interferer
is eectively disabled.
chanParams.adjacentUser1f if: The carrier frequency of the rst interfering
user after conversion to IF.
chanParams.adjacentUser1Chan: The time-invariant channel impulse response
of the rst interfering user.
chanParams.adjacentUser2Power: The relative power of the second interfering
user, specied in dB.
chanParams.adjacentUser2f if: The carrier frequency of the second interfering
user after conversion to IF.
chanParams.adjacentUser2Chan: The time-invariant channel impulse response
of the second interfering user.
chanParams.NBIfreq: The frequency of a narrowband interferer, after conversion
to IF.
chanParams.NBIPower: The relative power of the narrowband interferer, specied in dB.

458

Appendix H. The B 3IG Transmitter

rfParams.f s: The sampler frequency.

rfParams.T t: The nominal (i.e., expected) symbol period.
rfParams.T t err: The percentage error in the transmitters symbol period,
which is presumed unknown at the receiver.
rfParams.f if: The assigned intermediate carrier frequency (or frequencies).
rfParams.f if err: The percentage error in the transmitters local oscillator,
which is presumed unknown at the receiver.
rfParams.phaseNoiseVariance: The variance of the additive noise used to
model phase noise in the transmitter.
rfParams.SRRCLength: The length of the square-root raised cosine lter.
rfParams.SRRCrolloff: The rollo factor of the square-root raised cosine lter.

H.2.3

Overview of the BigTransmitter.m Code

While the B 3IG transmitter is a bit more sophisticated than the simple communications systems of Chapter 9, it still consists of the same basic operations. This
section provides a high-level outline of the code in BigTransmitter.m, which can
be compared with idsys.m of Chapter 9. Peeking inside the B 3IG transmitter
code, you will nd many familiar operations. In addition, acquainting yourself
with the inner workings of code will help prepare you for the receiver design
challenges that await. The commented sections in BigTransmitter.m point to
the code that implements the following operations:
Parse message, convert to bits, perform coding if requested, insert preamble, and
generate 4-PAM sequence
Create adjacent user signals
Add in main transmission, narrowband interference, adjacent users, and AWGN
Calculate the transmitters slightly inaccurate sample rate and IF
Upsample signal, perform pulse-shaping
Generate phase noise process
Mix signal to IF, while incorporating phase noise
Pass signal through channel (time-invariant case)
Pass signal through channel (time-varying case)
Perform downsampling and normalization (i.e., AGC)
Exercise H.1. Figure H.7 shows a typical received signal and its spectrum as
generated by B 3IG. The desired message is centered at 10 MHz and has an
absolute bandwidth of 1 MHz. As seen in the gure, there is an interfering
adjacent user centered at baseband with a 1 MHz absolute bandwidth. Assume
that the desired users signal was transmitted at a frequency of fRF = 300 MHz
and that the local oscillator shown in Figure H.5 operates at fLO = 90 MHz.
Assume that the preselect and IF lters are ideal bandpass lters from 250 to
350 MHz and 200 to 260 MHz, respectively. Complete the transmitter script
in BigEx2.m to generate a received signal identical to Figure H.7. Note that

The B 3IG Transmitter

459

Amplitude

Magnitude (x104)

0.004

0.008
(a) Seconds

0.012

Figure H.7 (a) received signal r[l]

and (b) its magnitude spectrum

R(f ).

8
6
4
2
0

0
1
(b) Frequency (x107)

three parameters have been left incomplete, as indicated by the question marks
in BigEx2.m. Determine the intermediate frequencies of the desired user and
adjacent interfering user, and the sampling frequency fs in order to generate the
signal in Figure H.7. Plot the spectrum of the signal you generate.

Exercise H.2. The script in BigEx3.m generates an ideal received signal much
like idsys.m from Chapter 9. This time, however, the transmitter sends signals
for two users using TDMA, where frameParams.userDataLength = 5. Mimic
the receiver code in BigIdeal.m to accurately decode the signal for each user.
Write your receiver code so that it stores the message for the rst user in a
variable called decoded1, and the message for the second user in a variable
called decoded2.

H.3

Notes on Debugging and Signal Measurement

In experimenting with the B 3IG transmitter, you may nd that it will be very
useful to look closely at intermediate signals the are inside the transmitter function. This can be accomplished in a variety of ways. Some possibilities are listed
here.

r Add more output arguments to BigTransmitter.m so that the variables are

available in the Matlab workspace.

r Use the debugging capabilities of Matlab, including breakpoints, step execution, etc.

r Add plotting commands within the transmitter code.

Index

(5, 2) block code, 330, 333, 334

(6, 4) block code, 337
(7, 3) block code, 336
16-QAM, 361
4-PAM, 5, 30, 153, 202, 205, 208, 212, 223,
232, 258, 293, 326, 342
4-QAM, 360
6-PAM, 307
adaptive components, 191
adaptive derotation, 386
adaptive element, 5556, 65, 114
Costas phase tracking, 208
CV clock recovery, 256
DD equalization, 288
DD for phase tracking, 212
DMA equalization, 291
LMS equalization, 285
LMS for AGC, 123
output power, 261
PLL for phase tracking, 203
SD for phase tracking, 198
simple AGC, 123
testing, 355
Aeneid, 314
AGC, 55, 120128, 166, 175
aliasing, 28, 51, 52, 99, 102
alphabet, 4, 35, 152, 211, 291, 323
AM
large carrier, 81, 419
suppressed carrier, 84, 419
analog vs digital, 11, 28, 30, 121
Anderson, J. B., 39, 377, 380, 402
angle formulas, 405
automatic gain control, see AGC
averaging, 122124, 199, 256, 261, 285,
442450
averaging and LPF, 442
bandlimited, 18, 101, 418
bandpass lter, see BPF
bandwidth, 8, 18, 85, 101, 102, 155, 230, 414
bandwidth vs data rate, 319

460

bandwidth vs SNR, 320

Barry, J. R., 383, 396, 402
baseband, 9, 247, 251, 271, 285
Baum, Frank, 310
Bello, P. A., 79
B 3IG transmitter, 187190, 215, 265, 354,
451459
binary arithmetic, 329
binary to text, 154
Bing, B., 356
Bingham, J. A. C., 356, 377, 392393, 402
bit error rate, 36
bit error vs symbol error, 310
bits denition, 307
bits to letters, 4, 153
bits to text, 4, 153
blind equalization, 289, 290
block codes, 314
BPF, 22, 24, 46, 60, 63, 76, 141, 151, 377
phase shift, 196, 197, 203
Buracchini, E., 356
Burrus, C. S., 151
capacity, see channel capacity
carrier
frequency, 17, 20, 22
recovery, 9, 33, 192221, 367
recovery, tuning, 351
Carroll, Lewis, 310
CD encoding, 339
CDMA, 22
cell phones, 17
center spike initialization, 292
Cervantes, M., 314
chain rule, 410
channel, 5, 11, 62, 164, 179, 221, 237, 260,
266, 270, 272
channel capacity, 37, 318323
CIRC encoder, 339
clock recovery, 33, 251269
decision-directed, 256
output power, 261
tuning, 351

Index

cluster variance, 174, 251, 254, 256

code division multiplexing, see CDMA
codeword, 333
coding, 4, 37, 303340
block, 314, 328
channel, 328
eciency, 324
instantaneous, 324
majority rules, 328
prex, 324
source, 323327, 339
variable length, 325
colored noise, 414
complex equalization, 282
complex-valued modulation, see modulation
component architecture, 15
compression
mp3, 327
uuencode, 327
zip, 327
computers and meaning, 310
constellation diagram, 174, 379
convolution, 23, 70, 72, 131, 142, 156, 158,
275, 408
correlation, 152, 157160, 172, 178, 245, 374,
426
correlation vs convolution, 160
Costas loop
PAM, 206
QAM, 367374
Costas, J. P, 225
Couch, L. W. III, 39
data rate, 8, 102
data rate vs bandwidth, 319
DD, see decision-directed
decision, 35, 174
directed equalization, 288
directed phase tracking, 211
hard, 35
soft, 36, 176
decoder, 328
delay spread, 270, 272
function, 65
discrete, 67
sifting property, 66, 131
spectrum, 67
demodulation, see modulation
demodulation via sampling, 103
dependence of English text, 311
derotation, 385
design methodology, 345
destructive interference, 339
dice, 307
dierence equation, 65, 430
dierential encoding, 375

461

digital radio, 1, 14, 342

discrete frequencies, 132
dispersion minimization, 290, 386
distortionless, 238
DMA, see dispersion minimization
Don Quixote, 314
Doppler eect, 25, 64
downconversion, 25
via sampling, 103
downsampling, 51, 109, 161, 173, 185
dual PLLs, 218
duality, 23, 407
dynamic range, 121
eciency, 324
electromagnetic transmission, 16
encoder, 328, 333
encoding a CD, 339
energy, 18, 409, 425
English
dependency, 311
frequency of letters, 311
random, 312
entropy, 315318, 324
envelope, 81, 416420
envelope detector, 416
equalization, 27, 34, 56, 270302, 387391
blind, 289, 290
complex, 282
dispersion minimization, 290
fractionally spaced, 283
initialization, 289, 292, 353
tuning, 352
error measures, 35
error surface, 118, 125, 200, 207, 213, 259,
261, 263
errors in transmission, 308
Eulers formulas, 404
exclusive OR, 329
eye diagram, 173, 181, 185, 231237, 241,
250, 435
fading, 10, 64, 127, 175
FDM, 22, 25, 29, 60
FFT, 18, 42, 131, 133141, 422424
frequency estimation, 194
of a sinusoid, 136
phase estimation, 195
vs DFT, 424
lter design, 76, 77, 130, 141, 146
lters, 23, 46, 62, 141, 275
nal value theorem, 117, 408, 410, 431
rpm, see Matlab rpm
xes, overview, 191
at fading, 175
Fourier transform, 18, 41, 131, 405408

462

Index

meaning of, 422

vs DFT, 421
fractionally spaced equalization, 283
frame, 4
Franks, L. E., 225
frequency, 16
carrier, 20
content, 19, 20, 40
discrete, 132
division multiplexing, see FDM
domain, 23, 40, 72, 132, 140, 421
intermediate, 29, 93, 130, 192
measuring, 18
of letters in English, 311
oset, 85, 88, 179, 184, 266
radio, 9
resolution, 140
response, 23, 65, 74, 133, 432
scale, 408
selectivity, 27
shift, 407
synchronization, 85
tracking, 216, 218
translation, 9, 17, 20, 21, 24, 54, 8097,
103, 192
freqz, see Matlab freqz

important message, 167

impulse response, 23, 65, 70, 74, 131, 141,
142, 270
impulse sampling, 100
independence
of English text, 311
independent events, 306
information, 37, 303
and complexity, 308
and uncertainty, 305
axioms, 306
denitions, 304
in a letter, 308
in digits of , 308
vs meaning, 310
initialization, center spike, 292
instantaneous code, 324, 325
instructor, to the, vii
integration layer, 341
interference, 10, 59
intermediate frequency, 29, 93, 192, 342
interpolation, 11, 110, 257, 262
intersymbol interference, see ISI
ISI, 34, 229230, 241, 247, 266, 270, 273,
387, 435
iteration, 447

gain of a linear system, 426

Galois elds, 339
generator matrix, 328
Gitlin, R. D., 397, 403
gong analysis, 139, 149
gradient, 34
gradient descent, 114
Gray code, 154, 361

Jablon, N. K., 394, 403

Jayant, N., 356
jitter, visualizing, 231
Johnson, C. R. Jr., 129, 302

Hamming
blip, 156, 169, 194, 228, 233
distance, 333
wide blip, 230
hard decision, 174
Haykin, S., 39, 56
HDTV, 102
header, 158, 344, 454
high-side injection, 93
highpass lter, see HPF
Hilbert transform, 407, 420
hill climbing, 115
HPF, 24, 46, 141
Human code, 37, 324
human hearing, 102
ideal
channel, 5
receiver, 4, 5, 13, 58, 165
transmitter, 165

Lathi, B. P., 39
least mean square algorithm, 285, 388
least squares, 277
Lee, E. A., 375, 386, 395, 402
Leibniz rule, 410
letters to bits, 4, 153
linear, 20, 23, 407, 422
linear block codes, 37
linear lters, 59, 65, 70, 130151
linear vs nonlinear codes, 335
LMS, see least mean squares
local minima, 118
logical AND, 329
lookfor, see Matlab lookfor
low-side injection, 93
lowpass lter, see LPF
LPF, 23, 141, 171
LPF and averaging, 442
M6 receiver, 342356
majority rules, 328, 332
marker, 162, 172, 178, 344, 454
matched lter, 242249
mathematical prerequisites, v

Index

Matlab
AGC, 124
averaging, 124
block code, 331
clock recovery, 258, 262
convolution, 70, 237
correlation, 158
DMA equalizer, 292
envelope, 416
equalizer
DD, 289
LMS, 286
LS, 278
error surface, 125, 213, 259
eye diagrams, 233
FFT, 135
lter, 47, 48, 70, 142, 237
rpm, vii, 47, 77, 146, 415
frequency response, 74
frequency tracking, 219
freqz, 143
help, 42
ideal receiver, 169
ideal transmitter, 168
interpolation, 112
lookfor, 42
matched lter, 244
mod, 329
noise, 413
phase tracking, 198, 204, 208, 212
plotspec, 42, 51
quantalph, 53, 161
rand vs randn, 414
random, 43
random sentences, 312
remez, 47
resample, 113
reshape, 233
sampling, 98, 108
source code, 326
spectrum of a pulse sequence, 228
timing recovery, 258, 262
toeplitz, 276
maximum phase, 430
maximum-length pseudo-noise sequence, 162
McClellan, J. H., 56, 151
mean, 413
Meyr, H., 268, 356, 394, 403
minimum distance, 333, 338
minimum phase, 430
Mitola, J. et al., 356
Mitra, S., 57
mixers, 5354
mixing, 81
mod, see Matlab mod
modular arithmetic, 329, 337

463

modulation, 9, 17, 20
complex-valued, 359
large carrier, 81
quadrature, 9093, 283, 417
quadrature amplitude, 357403
single sideband, 81
small carrier, 8490
vestigal sideband, 81
Morse code, 323
moving average, 442
mp3 compression, 327
multipath, 10, 27, 34, 61, 179, 182, 270,
272273, 435
mystery signal, 355
Nahin, P. J., 97
naive receiver, 12
Nevermind, 344
noise, 42, 77, 179
broadband, 60, 76, 180, 242, 412
colored, 414
in-band, 10, 60
narrowband, 61, 270, 412
out-of-band, 10, 60
simulating, 412415
spectrum of, 413
thermal, 60, 76
white, 245, 412
nonlinear, 20
nonlinear vs linear codes, 335
nonlinearities, 5253, 196, 197, 376, 416
nonreturn to zero, 236
number of particles in the Universe, 336
numerical approximation to derivative, 256,
262
Nyquist
frequency, 50
pulse, 237242, 247
rate, 101
sampling theorem, 28, 101
open eye, 185, 232, 266, 278, 287, 290, 435
open-eye measure, 436
Oppenheim, A. V., 56
optimization, 34, 114
oscillators, 4446, 192, 210
other users, 60, 270
output power, 251, 254
oversampling, 109, 156, 168, 194, 228, 236
PAM, 5
parity check, 328, 330
Parsevals theorem, 133, 408, 425
passband, 141
pedagogical method, iv
performance function, 114, 263, 293, 381

464

Index

Costas loop, 207

CV timing recovery, 256
DD carrier recovery, 211
DD equalization, 289
DMA equalization, 291
for QAM, 381383
LMS equalization, 284
LMS for AGC, 122
LS equalization, 276
output power, 261
PLL carrier recovery, 203
SD carrier recovery, 198
simple AGC, 123
period oset, 185
periodicity, 49, 101, 406
phase
ambiguity, 374
maximum, 430
minimum, 430
oset, 86, 87, 179, 183, 192
shift, 404
splitter, 366
synchronization, 86
phase tracking
analysis, 206
Costas loop, 206, 367, 376
decision-directed, 210
dual algorithm, 218
phase-locked loop, 55, 202, 375
squared dierence, 197
PLL, see phase tracking
Porat, B., 57
power, 18, 76, 84, 120, 242, 409, 413, 425
and correlation, 426
power spectral density, 242, 245, 409,
425427
PPM, 362
prex code, 324, 325
Proakis, J. G., 39, 379, 403
probability, use of, v
properties of discrete-time signals, 132
pulse amplitude modulation, see PAM
pulse shaping, 3032, 155, 160, 161, 172, 194,
226249, 251, 273
pulse train, 68, 99
pulse-matched ltering, 160, 172
pulse-phase modulation, see PPM
QAM, 357403
Q3 AM receiver, 397, 398
QPSK, 361, 377, 383
quadrature phase shift keying, see QPSK
quantization, 36, 161, 211
Qureshi, S. U. H., 302
radio

AM, 17, 21, 80

digital, 14, 342
FM, 17, 21
raised cosine pulse, 239, 240, 248, 406
rand, see Matlab random
random
numbers, 413
seed, 414
receive ltering, 160, 226249, 251
receiver
M6 , 342
Q3 AM , 397, 398
design, 345
ideal, 165
smart, 16
testing, 350, 355
reconstruction, 110
rectangular pulse, 8, 31, 72, 111, 238, 248,
406
redundancy, 37, 304, 308315
Reed, J. H., 356
ReedSolomon codes, 339
replicas, 23, 24, 100, 103
resample, see Matlab resample
reshape, see Matlab reshape
resolution in time vs frequency, 140, 423
Robbins, T., xiii
sampling, 11, 28, 33, 49, 98129, 251
for downconversion, 103
sampling theorem, 28
Schafer, R. W., 57
Schwarzs inequality, 243, 245, 410
scrambling, 162
seed for random numbers, 414
Sethares, W. A., 302
Shannon, C., 3, 37, 305, 308, 312, 340
sifting property of impulse, 410
signal, 40, 41, 131
Signal Processing First, v
signal-to-noise ratio, see SNR
simple average, 442
simulating noise, 412415
sinc function, 8, 31, 72, 111, 138, 236, 238,
248, 406
sine wave, 21
spectrum, 69
single-sideband modulation, 81
sketching frequency response, 432
SNR, 60, 76, 244
SNR vs bandwidth, 320
soft decision, 174
software-dened radio, 2, 14, 342, 356
source coding, 37
source recovery error, 251, 256, 261, 276, 285
source vs channel coding, 304

Index

sources of error, 348

spectrum, 41, 132, 135
function, 67
magnitude, 18, 23
of a pulse sequence, 227
phase, 18
sine wave, 69
square-law mixing, 88
square-root raised cosine, see SRRC
squared error, 35
squaring, 52, 195, 207
SRRC, 113, 236, 240, 248, 344, 406
steepest descent, 114
Steiglitz, K., 151
step
1 to build a radio, 1
2 basic components, 15
3 idealized receiver, 58
4 adaptive components, 191
5 putting it all together, 341
step function, 407
stepsize, 116, 123, 126, 199, 204, 209, 220,
257, 262, 287, 290, 292, 350, 353, 371,
447
Stonick, V., 57
stopband, 141
Stremler, F. G., 39
symbol error vs bit error, 310
symbol recovery error, 35, 271
symmetry, 132, 405, 407
synchronization, 7, 3334
carrier, 9, 192
frame, 7, 34, 154, 161, 177, 344, 353, 454
frequency, 85
phase, 86
unnecessary, 84
syndrome table, 330, 336, 338
system, 40, 131
tapped-delay line, 275
Taylor, F. J., 56
TDM, 22
temperature in Wisconsin, average, 443
testing the receiver, 350
text to binary, 154
text to bits, 4, 153
Through the Looking Glass, 310
time, 250
delay operator, 428
division multiplexing, see TDM
domain, 23, 40, 132, 140, 421
resolution, 140
shift, 408, 410, 428
timing, 33
jitter, 7
oset, 6, 179, 185

recovery, 251269, 380384

Toeplitz matrix, 276, 278
tracking, 124
trade-os, 350
training sequence, 273, 284, 344, 454
transfer function, 74, 133, 430
transition
band, 142
probabilities, 313
transmitter
B 3IG, 187, 451
design, 343
ideal, 165
transpose, 411
Treichler, J. R., 394, 403
trial and error, 350
trigonometric identities, 404
tuning the receiver, 350
TV, 17
uniform sampling, 50
unit circle, 430
unit delay, 429
variable length code, 323
variance, 413
Vergil, 314
Verne, Jules, 314
vestigal sideband modulation, 81
wavelength, 16
what if?, 3, 167, 178
whatever, 344
Widrow, B., 129
wiener, 168
Wizard of Oz, 310, 325, 327
zip compression, 327
Z-transforms, 409, 428434

465