Sie sind auf Seite 1von 109

Queensland University of Technology

Signals in Linear Systems


Boualem Boashash
Signal Processing Research Centre, QUT
Queensland University of Technology
GPO Box 2434, Brisbane, QLD 4001

Copyright: B. Boashash

Last revised: March 1999.


Published by
Queensland University of Technology
GPO Box 2434, Brisbane, Queensland 4001
ISBN 1 86435 436 4

Copyright c B. Boashash, Signal Processing Research Centre, QUT 1999


First Published in September 1996

Printed in Brisbane

National Library of Australia


Cataloguing-in-Publication Data

Boashash, Boualem
Signals and linear systems.
Bibliography
ISBN 1 86435 436 4
1. Signal processing. 2. System analysis. 3. Linear time
invariant systems. I. Boashash, Boualem. II. Queensland
University of Technology. Signal Processing Research Centre.
III. Title.
621.3822
Contents
Preface vii
1 Introduction to Signals and Systems 1
1.1 The Basic Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.2 Signal Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.3 System Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2 System Analysis in the Time Domain 15
2.1 The Convolution Integral for Linear Shift-invariant Systems . . . . . . . . . . . . 15
2.2 Properties of the Convolution Integral . . . . . . . . . . . . . . . . . . . . . . . . 17
2.3 Step Response of a Linear System . . . . . . . . . . . . . . . . . . . . . . . . . . 18
2.4 Stability of Linear Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
3 Spectral Representation of Signals 23
3.1 The Fourier Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
3.1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
3.1.2 Trigonometric Fourier Series Representations for Periodic Signals . . . . . 24
3.1.3 The Exponential Fourier Series . . . . . . . . . . . . . . . . . . . . . . . . 25
3.1.4 Properties of the Fourier Series Coecients . . . . . . . . . . . . . . . . . 25
3.1.5 Representation of the Fourier Coecients in the Frequency Domain . . . 28
3.1.6 Systems with Periodic Inputs . . . . . . . . . . . . . . . . . . . . . . . . . 30
3.1.7 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
3.2 The Fourier Transform . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
3.2.1 Denition of the Fourier Transform . . . . . . . . . . . . . . . . . . . . . . 32
3.2.2 Convergence of the Fourier Transform . . . . . . . . . . . . . . . . . . . . 33
3.2.3 Fourier Transforms in the Limit . . . . . . . . . . . . . . . . . . . . . . . . 34
3.2.4 Energy Spectral Density . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
3.2.5 Properties of the Fourier Transform . . . . . . . . . . . . . . . . . . . . . 36
3.3 The Laplace Transform . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
3.3.1 Existence of the Laplace transform . . . . . . . . . . . . . . . . . . . . . . 39
3.3.2 Relation Between Fourier Transform and Laplace Transform . . . . . . . 41
3.3.3 Conditions for the Realisability of a Linear System or Filter . . . . . . . . 42
3.3.4 Frequency Response of Transfer Function using the Poles and Zeros of the
Laplace Transform . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
3.3.5 Poles and Zeros . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
3.4 Time{Varying Spectra . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
3.4.1 The Hilbert Transform . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
3.4.2 Analytic signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
i
ii CONTENTS
3.4.3 Time{Frequency Relationships . . . . . . . . . . . . . . . . . . . . . . . . 48
3.4.4 Time{Frequency Distributions . . . . . . . . . . . . . . . . . . . . . . . . 50
4 System Analysis in the Frequency Domain 53
4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
4.2 An Example of Frequency Domain System Analysis . . . . . . . . . . . . . . . . 53
4.3 Transfer Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
4.4 Steady{State System Response . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
4.5 Ideal Filters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
4.6 Fourier Transforms of Periodic Signals and Sampling . . . . . . . . . . . . . . . . 59
5 Introduction to Discrete{time Signal Processing 63
5.1 Analog to Digital Conversion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
5.1.1 Sampling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
5.1.2 Periodic Sampling: a tutorial . . . . . . . . . . . . . . . . . . . . . . . . . 67
5.1.3 Reconstruction of Band{Limited Signals . . . . . . . . . . . . . . . . . . . 70
5.1.4 Quantising and Encoding . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
5.2 Digital Processing of Continuous-Time Signals Using the DTFT . . . . . . . . . . 77
5.2.1 Discrete{time Processing of Continuous{time Signals. . . . . . . . . . . . 78
5.2.2 The Discrete{Time Fourier Transform . . . . . . . . . . . . . . . . . . . . 79
5.2.3 Convergence of the DTFT . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
5.2.4 Properties of the DTFT. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
5.3 The z{Transform . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
5.3.1 Denition of the z{Transform . . . . . . . . . . . . . . . . . . . . . . . . . 86
5.3.2 Properties of the z{Transform . . . . . . . . . . . . . . . . . . . . . . . . . 91
5.3.3 Inverse z{Transform . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
5.4 Digital Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
5.4.1 Dierence Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
5.4.2 Discrete Convolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
Bibliography 98

B. Boashash, Signals in Linear Systems


List of Figures
1.1 Representation of a system. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Time domain signal processing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.3 Frequency domain signal processing. . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.4 Square pulse. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.5 Example of a continuous-time signal. . . . . . . . . . . . . . . . . . . . . . . . . 4
1.6 An example of a discrete-time or sample-data signal. . . . . . . . . . . . . . . . . 4
1.7 Examples of aperiodic (a) and periodic, x(t) = sin(2t=T0 ) (b) signals. . . . . . . 5
1.8 Representation of z (t) in the frequency domain. . . . . . . . . . . . . . . . . . . . 5
1.9 The unit step function. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.10 The unit ramp function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.11 The unit impulse function. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.12 Square pulse approximate for the unit impulse function. . . . . . . . . . . . . . . 8
1.13 Plot of the function p1 (t): . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.14 Multiple-input multiple-output system. . . . . . . . . . . . . . . . . . . . . . . . . 11
1.15 Input and output signals in a time-invariant causal system. . . . . . . . . . . . . 12
1.16 System of second order. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.1 Linear shift-invariant system. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.2 The plot of x(), h(), h(;), and h(t ; ) as a function of . . . . . . . . . . . 20
2.3 Convolution of x(t) and h(t). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
2.4 Linear shift-invariant system. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
3.1 Periodic signal x(t) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
3.2 Allowing period T0 to increase to obtain the non-periodic signal. . . . . . . . . . 32
3.3 Triangular signal and its derivatives . . . . . . . . . . . . . . . . . . . . . . . . . 40
3.4 Region of convergence <fsg > ;. . . . . . . . . . . . . . . . . . . . . . . . . . . 41
3.5 System of second order. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
3.6 System of second order. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
3.7 Spectrum of real information signal. . . . . . . . . . . . . . . . . . . . . . . . . . 48
3.8 Spectrum of modulated information signal. . . . . . . . . . . . . . . . . . . . . . 48
3.9 Spectrum of real information signal. . . . . . . . . . . . . . . . . . . . . . . . . . 49
3.10 ..................................... . . . . . . . . . 49
4.1 Band-pass ltering of desired signal. . . . . . . . . . . . . . . . . . . . . . . . . . 53
4.2 Amplitude response of a low-pass lter. . . . . . . . . . . . . . . . . . . . . . . . 55
4.3 Phase response of a low-pass lter. . . . . . . . . . . . . . . . . . . . . . . . . . . 55
4.4 Time and frequency domain plots. . . . . . . . . . . . . . . . . . . . . . . . . . . 57
4.5 Transfer function of an ideal bandpass lter. . . . . . . . . . . . . . . . . . . . . . 59
4.6 Transfer function of an ideal high{pass lter. . . . . . . . . . . . . . . . . . . . . 59
iii
iv LIST OF FIGURES
4.7 Periodic triangular signal. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
5.1 ANALOG{DIGITAL CONVERTER . . . . . . . . . . . . . . . . . . . . . . . . . 63
5.2 Sampling of a continuous signal . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
5.3 A typical sampling process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
5.4 The waveform p(t). Ideally, p(t) should be a periodic set of delta functions . . . . 65
5.5 Exact recovery of a continuous{time signal from its samples using an ideal low{
pass lter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
5.6 Continuous-Time to digital converter. . . . . . . . . . . . . . . . . . . . . . . . . 68
5.7 Eect in the frequency domain of sampling in the time domain: (a) Spectrum of
the continuous-time signal, (b) Spectrum of the sampling function, (c) Spectrum
of the sampled signal with fs > 2fB , (d) Spectrum of the sampled signal with
fc < 2fB . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
5.8 Ideal reconstruction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
5.9 Sampling a single sinusoid at 70 and 140 Hz . . . . . . . . . . . . . . . . . . . . . 72
5.10 Sampling of a non{periodic signal at 70 and 140 Hz . . . . . . . . . . . . . . . . 73
5.11 Quantising and encoding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
5.12 Transfer characteristic (with null error) . . . . . . . . . . . . . . . . . . . . . . . 74
5.13 Transfer characteristic (without null error) . . . . . . . . . . . . . . . . . . . . . . 75
5.14 Calculation of quantising error . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
5.15 SNR versus word length . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
5.16 Discrete-time processing of continuous-time signals. . . . . . . . . . . . . . . . . . 78
5.17 Continuous-Time Filter. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
5.18 LSI system excited by x(n) = ej 2fn . . . . . . . . . . . . . . . . . . . . . . . . . . 80
5.19 Unit circle in the complex z -plane. . . . . . . . . . . . . . . . . . . . . . . . . . . 87
5.20 The unit pulse. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
5.21 The sequence p(n). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
5.22 Discrete Time Signal. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
5.23 The unit step sequence. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
5.24 Exponential Sequence with 0 <  < 1. . . . . . . . . . . . . . . . . . . . . . . . . 90
5.25 Region of Convergence ROC: Rx; < jz j < Rx+. . . . . . . . . . . . . . . . . . . 91
5.26 Linear digital signal processor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
5.27 Convolution operation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97

B. Boashash, Signals in Linear Systems


List of Tables
3.1 Some selected Fourier transform pairs. . . . . . . . . . . . . . . . . . . . . . . . . 35
3.2 Some selected unilateral Laplace transform pairs. . . . . . . . . . . . . . . . . . . 51
3.3 Some selected properties of the Laplace transform. . . . . . . . . . . . . . . . . . 52
5.1 Discrete{Time Fourier transform theorems. . . . . . . . . . . . . . . . . . . . . . 84
5.2 Discrete{Time Fourier transform pairs. . . . . . . . . . . . . . . . . . . . . . . . 85

v
vi LIST OF TABLES

B. Boashash, Signals in Linear Systems


Preface
1 The topic
The eld of Signal Processing has grown so much in the last three decades that it is now a
major industry in the world. The magazine of the IEEE, the largest worldwide professional
association in Electrical Engineering allocates most of its pages to the advertisement of new
Signal Processing products.
In universities and schools of Electrical Engineering, Signal Processing has become one of the
major streams. The unit 'Signals in Linear Systems' is the rst major Signal Processing subject
in the School of Electrical and Electronic Systems Engineering, QUT. This subject is a classical
subject, which has been taught in universities for many years - in Electrical Engineering as well
as in physics. The material can also be found in Mechanical Engineering, as well as Mathematics
in a dierent form. `Signals in Linear Systems' refers specically to the study of signals, the
study of linear systems and the interaction between signals and linear systems. An example of
a signal is voice and an example of a linear system is the telephone line. A common problem is
to remove the inuence of noise introduced during the interaction between signals and systems
through the design and use of suitable systems, which perform a ltering operation.

2 The material
The material contained in this document has been developed as lecture notes over a period of
more than 10 years. I wrote the rst draft in 1984 at the University of Queensland using as a
basis a course developed at the University of Lyon, where I lectured in 1982{3. I then further
adapted the notes with the aim of establishing a strong link between theory and practice,
while keeping the highest possible standards, logic and rigour in the text. I made this possible
by emphasising the importance of concepts and the philosophy of signals processing, and the
inspiration provided by early applications and developments in the eld of geophysical signal
processing, where some of the early developments in DSP took place. (I witnessed rst hand the
power of DSP in my 3 year long career in industry, as a research engineer in the seismic signal
processing division of Elf{Aquitaine, one of the biggest European oil companies.)
Further, the process of learning was facilitated by the use of suitable cross{references within
the text to previously acquired knowledge, and generally speaking, the presentation of the
material as a natural extension of basic principles already mastered by the students.
The lecture notes were revised every year since 1984. A major revision occurred in 1991
when I moved to the Queensland University of Technology, with the task of establishing a
strong undergraduate signal processing program as well as a research centre of international
standard. Since 1991, the material continued to be upgraded and taught under my supervi-
sion. A complementary booklet containing the exercises and examples covered in tutorials is in
preparation.
vii
viii Preface
This subject is the prerequisite for the subject \Digital Signal Processing" (DSP). I hope
that this eort will help students towards mastering these important concepts and techniques,
and I look forward to receiving any suggestions for further improvement of this material.
The material contained in these notes comes straight from the original lecture notes of the
author, except for some improvements in the quality of gures and some examples that have
been added.
Parts of this manuscript appeared in an earlier joint publication with Dr. Zoubir (B.
Boashash and A. Zoubir, Signals in Linear Systems, QUT, 1996) whom I wish to thank.
The author also acknowledges the valuable help provided by Dr B. Senadji who checked the
text and by the various research assistants who assisted me in the revision of the notes.

3 The course organisation


3.1 Organisation and Contents of the Book
The book is organised in ve major chapters.
 Chapter 1: introduces the subject and describes existing models for signals and sys-
tems
 Chapter 2: presents time domain signal and system analysis methods
 Chapter 3: is a study of the important `Transforms in Signal Processing' such as, the
most important `Fourier Transform' and other related transforms
 Chapter 4: presents frequency domain signal and system analysis methods and
 Chapter 5: deals with the most important concepts of sampling and quantizing and
introduces the representation and analysis of digital signals.
3.2 Teaching Organisation and Assessment is as follows:
Lec Tut Total
Hours per week: 2.0 1.0 3.0
Total hours: 24 12 36
Semester Examination: 85% (Minimum 40%)
Assignments: 15% (Minimum 40%)
Assignments will be handed out throughout the semester with selected problems being
submitted for assessment. Answers to assignments will be provided at appropriate times.

Dr Boualem Boashash Brisbane, Australia


Professor of Signal Processing date: March 1999
Director, Signal Processing Research Centre

0 Warning and disclaimer: Due to the ongoing process of revision, this draft document may contain typos
and errors due to the editing process. These will be corrected during the lectures and in the next version.

B. Boashash, Signals in Linear Systems


Chapter 1

Introduction to Signals and Systems


\Signal Processing" is a new eld which has proved useful in many diverse elds. The methods
and techniques of Signal Processing are used to:
 characterise physical systems and signals
 determine the interaction of systems and signals
 extract information from the signal
The material presented in this book deals with systems and the interaction of signals in such
systems. A system is dened as a combination and interconnection of several components to
perform a desired task. Such a task might be the amplication of music to a desired level or the
transmission of information from one town to another. We will be concerned with linear systems,
because many systems of engineering interest are closely approximated by linear systems and
very powerful techniques exist for analysing them.
Signals. A signal is any function of time which represents a physical variable of interest as-
sociated with a system. In electrical systems, signals usually represent currents and voltages,
whereas in mechanical systems, they might represent forces and velocities. In hydraulic systems
they could be pressure and ow rate in optical systems: light intensity and ux in thermal
systems: temperature and heat ow.
Systems. A system represents any process that results in the transformation of signals in-
cluding basic operations such as additions, multiplications and integrations. It maps an input
signal into an output signal. Systems may be described using a block diagram. For example,
consider the system in Figure 1.1 with the input x(t) and the output y(t).

x(t) - System - y(t)

Figure 1.1: Representation of a system.


The input signal x(t) and the output signal y(t) are shown as arrows, and the system itself
as a block. We consider single input{single output (SISO) systems, i.e. x(t) and y(t) are scalars.
The main possibilities are: (i.e. the main possible practical problems are:)
1
2 Chapter 1. Introduction to Signals and Systems
1. Input and system are known, output is unknown.
2. Input and output are known, system is unknown.
3. System and output are known, input is unknown.

1.1 The Basic Techniques in Signal and System Analysis


There are several methods of system analysis and there are also several methods of representing
and analysing signals. They are not all convenient in every particular situation. We will point
out useful applications of each technique.
A system can be studied from:
 A microscopic point of view
It deals with the full description of the system, including the interactions of all various
internal components. This point of view is very dicult or even impossible due to the
large complexity in any mathematical description of the system.
 A macroscopic point of view
We consider the system in terms of subsystems and components. Each part is described
not precisely, but in sucient detail so that the full system operation can be predicted.
We model the system in terms of component behaviour. The behaviour is dened by
description of the expression of the response of each component (or subsystem) to a forcing
function (impulse response, step response, transfer function).
Deterministic and non-deterministic signals. Deterministic signals can be modelled as
completely specied functions of time. For example, the signal
x(t) = A cos(2f0 t + )
;1 < t < 1 (1.1)
where A, f0 and are constants, is a deterministic signal. At every time t, x(t) is known
precisely. A second class of signals are random or non-deterministic signals. They are signals
taking on random values at any given time instant and must be modelled probabilistically. This
class of signals will be discussed in another text.

x(t) - System, h(t) - y(t) = x(t)  h(t)

Figure 1.2: Time domain signal processing.


Time domain signal processing. The system is characterised by the impulse response, h(t),
as shown in Figure 1.2. The relation between x(t) and y(t) is given by the convolution integral
Z1
y(t) = x( )h(t ; )d = x(t)  h(t) (1.2)
;1

B. Boashash, Signals in Linear Systems


1.2. Signal Models 3
when the system is linear and shift{invariant (LSI).
In the time domain, the correlation function is used to characterise the relationship between
two signals representing 2 processes. Applications include:
- detection of known signals in a noisy environment (sonar, radar)
- detection of intelligence in a random signal, e.g. the basic method of analysing signals
received from the universe is to "autocorrelate" them. If they convey the intelligence,
there should be some predictive information in the message.
Frequency domain signal analysis. A common technique is to transform a signal into
its equivalent form in the frequency domain. We can obtain its spectrum using the Fourier
transform as a tool.

X (f ) - System, H (f ) - Y (f ) = X (f )  H (f )

Figure 1.3: Frequency domain signal processing.


The function H (f ), called the frequency response or the transfer function, is given by
Z1
H (f ) = h(t)e;j 2ft dt:
;1
Here we particularly choose the Fourier transform because
ej!t = ej2ft = cos 2ft + j sin 2ft
are the eigenfunctions1 of a linear system and they represent the natural choice when one deals
with signals in linear systems. Moreover, the Fourier transform is not only a theoretical concept
but also a reality, e.g. in optics. Optical systems can perform linear transformations in two
dimensions - using a lens we can produce a two-dimensional (2D) Fourier transform.

1.2 Signal Models


Continuous{Time Signals. A continuous{time signal is a signal represented by a function
of a continuous{time variable an example is the function rect(t=T ), depicted in Figure 1.4. This
doesn't imply that the signal is a mathematically continuous function. Examples of continuous-
time signals are speech, image, seismic and radar signals. An example of a continuous time
signal is shown in Figure 1.5.
Discrete{Time Signals. In some systems, the signals are represented only at discrete values
of time. Between these instants, the value of the signal may be zero, undened, or of no interest.
Figure 1.6 is an example of a discrete{time or sample{data signal. Very often, the intervals
between signal values are the same, but they need not be. Example of a discrete-time signal:
monthly payment of a home mortgage.
1 An eigenfunction of a system is an input signal that produces an output that diers from the input by a
constant multiplicative factor. This factor is called the eigenvalue of the system.

B. Boashash, Signals in Linear Systems


4 Chapter 1. Introduction to Signals and Systems
rect(t=T )
6
1

-
;T
2
T
2 t
Figure 1.4: Square pulse.

x(t)

Figure 1.5: Example of a continuous-time signal.

Quantised, Discrete-time and Digital Signals. A quantised signal is one whose values
may assume only a countable number of values, or levels, but the changes from level to level
may occur at any time. A discrete-time signal is one discretised in time. A quantised and
discrete-time signal is referred to as digital signal (discretised in both time and amplitude).
Examples of digital signals are computer data.
Periodic and Aperiodic Signals. A signal x(t) is periodic if and only if there exists a
constant T0 > 0 such that
x(t + T0 ) = x(t)
;1 < t < 1

where T0 is called the period. Any signal that does not satisfy this property is called aperiodic.

x(t)

t1 t2 t3 t4 t5 t

Figure 1.6: An example of a discrete-time or sample-data signal.

B. Boashash, Signals in Linear Systems


1.2. Signal Models 5
x(t) x(t)

t
Ae

t T0 t

(a) (b)

Figure 1.7: Examples of aperiodic (a) and periodic, x(t) = sin(2t=T0 ) (b) signals.

Complex Signals. It is often convenient to represent real signals in terms of complex quan-
tities. For example, given the real, sinusoidal signal, x(t) = A cos(2fo t + ), we refer to the
complex signal
z (t) = A ej(2f0 t+)
x(t) is obtained from z (t) through
x(t) = <fz(t)g:
These expressions for x(t) are called time{domain representations. An alternative represen-
tation for x(t) is provided in the frequency domain: Since z (t) = A ej (2f0 t+) is completely
specied by the real numbers A and for a given value of f0 , this alternative frequency{domain
representation can take the form of two plots, one showing the amplitude A as a function of
frequency f , and the other as a function of f as depicted in Figure 1.8.
jz(t)j (z(t))
6 6
jAj + 2f0 t

- f - f
0 f0 0 f0
Figure 1.8: Representation of z (t) in the frequency domain.

Singularity Functions. An important class of signals very much used in signal processing
are the singularity functions.
The unit step function.
The unit step function, u(t) is dened as

u(t) = 0
t<0
1
t0
B. Boashash, Signals in Linear Systems
6 Chapter 1. Introduction to Signals and Systems
6u(t)
1
- t
0
Figure 1.9: The unit step function.

In terms of distribution theory, one usually assigns u(t) = 21 , for t = 0.


Example 1.2.1 A mathematical representation of the following time function is given by
6x(t)

3
2
1
-
t
;11 3 5 7
;2
x(t) = 4u(t) ; 6u(t ; 2) + 3u(t ; 4) + 2u(t ; 5)
The unit ramp function.
The unit ramp function r(t) is dened as

r(t) = t
t0
0
t<0

6r(t) 
 1
 j
 - t
0 1
Figure 1.10: The unit ramp function

The unit impulse function.


B. Boashash, Signals in Linear Systems
1.2. Signal Models 7
The unit impulse function or delta function (Dirac function), (t), has the properties:
(t) = 0
t 6= 0
Z1
(t)dt = 1
;1
Note that
Zt
( )d = u(t)
;1

and
Zt
u( )d = r(t):
;1

6 (t)
6
- t
0
Figure 1.11: The unit impulse function.

The motivation for dening a function with such properties is due to the need for representing
phenomena that happen in time intervals that are short when compared with the resolution
capability of any measuring apparatus used, but which produce an almost instantaneous change
in a measured quantity.
When dening the Dirac or delta function by means of distributions, it is possible to consider
the derivative of a delta function, even though it does not exist in the usual sense as a function.
As (t) is not a usual function, we dene it by considering the limit of a conventional function
as some parameter approaches zero. For example, consider the signal

(t) = 1
rect( t ) = jtj ""
"

1
" " " 0
jtj > 2
2

No matter how small " is, " (t) always has unit area. If " ! 0, the height of " (t) ! 1 such
that unity area lies between the function and the t{axis.
There exist other functions which converge to the impulse function as " ! 0. These include
for example
 1 t 2
p1 (t) = " t sin "
which is depicted in Figure 1.13.
B. Boashash, Signals in Linear Systems
8 Chapter 1. Introduction to Signals and Systems
6
" (t)
1
"

-
;" 2
"
2 t
Figure 1.12: Square pulse approximate for the unit impulse function.
2

1.8

1.6
epsilon=1.0
1.4 epsilon=0.5

1.2
p1(t)

0.8

0.6

0.4

0.2

0
5 4 3 2 1 0 1 2 3 4 5
t

Figure 1.13: Plot of the function p1 (t):

Denition 1.2.1 (t) is a function dened by an integral function which assigns the value x(0)
to any function x(t) continuous at the origin
Z1
x(t) (t)dt = x(0)
;1
The more general case of the above equation is
Z1
x(t) (t ; t0 )dt = x(t0 )
;1
Properties of (t).
Assuming that x(t) is continuous at t = t0 , we have:
1. (at) = 1 (t)
jaj
2. (;t) = (t)
Z1
3. x(t) (t ; t0 ) dt = x(t0 )
;1
Z1
4. x() (t ; ) d = x(t) important (convolution)
;1
5. x(t) (t ; t0 ) = x(t0 ) (t ; t0 )
Z1 Z1
6.
;1
0 (t ; )x() d = ;1
(t ; )x0 () d,

B. Boashash, Signals in Linear Systems


1.2. Signal Models 9
where 0 (t) = dtd (t), and x0 (t) = dtd x(t). The derivative of the impulse function, denoted by
0 (t), is dened by
Z t2
x(t) 0 (t ; t0 )dt = ;x0(t0 )
t1 < t0 < t2
t1
provided x(t) possesses a derivative x0 (t) at t = t0 . Then
Z t2 Z t2
t1
x(t) 0 (t ; t )dt
0 =
t1
x(t)d 0 (t ; t0)]
Z t2
= x(t) (t ; t0 ) jt1 ; x0 (t) (t ; t0 )dt
t 2
t1
= 0 ; 0 ; x (t0 ):
0

Energy and Power Signals.


Denition 1.2.2 For an arbitrary signal x(t), the total energy (normalised to unit resistance)
is given by
ZT
E = Tlim
!1 ;T
jx(t)j dt
2
Joules]

and the average power normalised to unit resistance is dened as


1 ZT
P = Tlim
!1 2T ;T
jx(t)j dt2
Watts]

We dene the two following classes of signals


1. x(t) is an energy signal if and only if 0 < E < 1, so that P = 0 (x(t) is a transient).
2. x(t) is a power signal if and only if 0 < P < 1, thus implying that E = 1.
Example 1.2.2 Consider the following signal
x1 (t) = A e;t u(t) >0 A2R
ZT
E = Tlim
!1 ;T
jA e;t u(t)j2dt
ZT
= Tlim
!1 0
jAj2 e;2tdt
= Tlim jAj2 e;2T ; 1
!1 ;2
= jA
2
j2 :
Thus, the signal x1 (t) is an energy signal.

B. Boashash, Signals in Linear Systems


10 Chapter 1. Introduction to Signals and Systems
Example 1.2.3 Consider the following signal
x2 (t) = A u(t)
1 ZT
P = Tlim
!1 2T ;T
jA u(t)j2 dt
1 ZT
= Tlim
!1 2T 0
jAj2dt
= Tlim 1 jAj2 T
!1 2T
= jA2j :
2

Thus, the signal x2 (t) is a power signal.


Average Power of a Periodic Signal. The average power of a periodic signal is dened as
1 Z t0 +T0
P=T
0 t0
jx(t)j2dt

where T0 is the period of x(t).


Example 1.2.4 The power of x(t) = A cos(!0t + ) is given by
1 Z t0 +T0
P = T A2 cos2 (!0 t + )dt
0t0
Z t0 +T0 1
= AT
2

0 t0 2 (1 + cos 2(!0 t + )) dt
A 2 A2 Z t0 +T0
t
= 2T t]t0 + 2T
0 +T0
cos 2(!0 t + )dt
0 t0
0
| {z }
=0

= A
2

2
1.3 System Models
We want to look at ways to represent the eects of systems on signals, i.e. we want to design
appropriate system models that represent correctly the interaction of signals and systems and
the relationship of causes and eects for that system.
Representations for Systems. We will express the dependence of the system output on the
input symbolically as
y(t) = T x(t)]
which is read: y(t) is the response of T to x(t).
The operator T identies the system and species the operation to be performed on x(t) to
produce y(t). For a multiple{input, multiple{ output system, x(t) and y(t) are vectors and T is
a matrix.
B. Boashash, Signals in Linear Systems
1.3. System Models 11
x1(t) - - y1 (t)

x2(t) - - y2 (t)

..
. System T ..
.

xp(t) - - yq (t)

Figure 1.14: Multiple-input multiple-output system.

In Figure 1.14 a system model is shown with p inputs fx1


x2
: : :
xp g and q outputs
fy
y
: : :
yp g.
1 2 The inputs are sometimes referred to as the excitations, causes or stimuli.
The outputs are sometimes referred to as responses or eects.
An example of a system input{output relation is the integral representation that will be
studied later for linear systems:
Z1 Z1
y (t ) = h() x(t ; ) d = x() h(t ; ) d

;1 ;1
which is known as the superposition integral or convolution. The function h() is called the
impulse response of the system and h() is the response of the system to a Dirac impulse applied
at t = 0, i.e. T  (t)] = h(t).
Properties of Systems
 Continuous{time and Discrete{time systems
If the signals processed by a system are continuous{time signals, the system itself is referred
to as a continuous time system. If, on the other hand, the system processes signals that
exist only at discrete times, it is called a discrete{time system. If the signals are quantised
to a nite number of levels, the system is called a quantised system. If the quantised system
is also a discrete{time system, then it is called a digital system.
 Fixed and time{varying systems

A system is time{invariant, or xed or shift{invariant if its input{output relationship does


not change with time. Otherwise, it is called a time{varying system.
Denition 1.3.1 A system is xed if and only if
8x(t)
8 T x(t ; )] = y(t ; )
 Causal and non-causal systems
A system is causal if its response to an input does not depend on future values of that
input.
B. Boashash, Signals in Linear Systems
12 Chapter 1. Introduction to Signals and Systems
x(t) y(t)

x(t) y(t)
System

t0 t1 t t0 t1 t

Figure 1.15: Input and output signals in a time-invariant causal system.

Denition 1.3.2 A continuous{time system is causal if and only if the condition x1(t) =
x2 (t) for t t0 implies the condition T x1 (t)] = T x2 (t)] for t t0 for any t0 , x1 (t) and
x2 (t).
 Instantaneous and dynamic systems
A system for which the output is a function of the input at the present time only is said
to be instantaneous (or memoryless).

A dynamic system (or a system with memory) is one whose output depends on past or
future values of the input in addition to the present time. If the system is also causal, the
output of a dynamic system depends only on present and past values of the input. An
example of a causal and time-invariant system is given in Figure 1.15.
Example 1.3.1 A resistor R is an instantaneous system,
i(t) R

v(t)

v(t) = R i(t)
Example 1.3.2 For an inductor L
i(t) L

v(t)

1 Zt
i(t) = L v(x)dx:
;1
This relation shows clearly that i(t) depends on past values of v(t). Thus an inductor L is
a dynamic causal system.
B. Boashash, Signals in Linear Systems
1.3. System Models 13
 Linear and nonlinear systems
A linear system is a system for which the theorem of superposition is valid that is, if x1 (t)
and x2 (t) are two inputs, we must have for any arbitrary constants 1 and 2 and any t
T  x (t) +  x (t)]
1 1 2 2 = 1 T x1 (t)] + 2 T x2 (t)]
= 1 y1 (t) + 2 y2 (t):
The property
T x(t)] = T x(t)]
is called homogeneity and
T x (t) + x (t)] = T x (t)] + T x (t)]
1 2 1 2

is referred to additivity.
Example 1.3.3 (System dierential equation) Let a system be given in Figure 1.16. A
R

L
vi (t) C vo (t)

Figure 1.16: System of second order.


relationship between vi (t) and vo (t) can be obtained by

vi (t) = RC dvdto (t) + LC d dt


vo (t) + v (t)
2
2 o

or equivalently

LC d dt
2 v (t)
o + RC dvo (t) + v (t) = v (t)
2 dt o i

This system is linear and time-invariant.

B. Boashash, Signals in Linear Systems


14 Chapter 1. Introduction to Signals and Systems

B. Boashash, Signals in Linear Systems


Chapter 2

System Analysis in the Time Domain


2.1 The Convolution Integral for Linear Shift-invariant Systems
The system analysis problem is: given a system and an input what is the output? There are

x(t) - System T - ?

Figure 2.1: Linear shift-invariant system.

two dierent approaches to solve this problem that make use of signal theory1 .

1. Carry out a solution in the time domain using the convolution integral.

2. Use frequency-domain analysis by means of the Fourier Transform or the Laplace Trans-
form.

These two methods are of common use, and will be discussed in this chapter for LSI systems.
In this chapter, we will examine the rst method. We will concentrate on linear systems
because many of the physical systems can be very well modelled as linear systems. A very
important operation in time-domain analysis techniques is the convolution integral.

Denition 2.1.1 The impulse response of a system is its response to a unit impulse at time
t = 0 with all initial conditions of the systems zero.
1 We coul also describe the interaction of the system and the signals by a dierential equation and solve it.
This is a dicult task in practice and is better replaced by signal processing theory.

15
16 Chapter 2. System Analysis in the Time Domain
(t) ;! h(t) read h(t) is the response of (t)
because
(t ; ) ;! h(t ; ) for all  2 R shift{invariance
Z 1 ) (t ; )
x( ;! x()h(Zt ;1 ) for all x(),  2 R homogeneity
x() (t ; )d ;! x()h(t ; )d additivity
;1 Z;11
which is equivalent to x(t) ;! x()h(t ; )d convolution integral Thus
;1
Z1 Z1
y(t) = x()h(t ; )d = h()x(t ; )d
;1 ;1
Denition 2.1.2 If x(t) and h(t) are two signals, the convolution of x(t) with h(t) is a new
signal y(t) given by the operation
Z1
y(t) = x()h(t ; )d
;1 < t < 1
;1
Alternatively, y(t) can be found by convolving h(t) with x(t) which is expressed as
Z1
y(t) = h()x(t ; )d
;1 < t < 1
;1
A useful symbolic notation is very often employed
y(t) = x(t)  h(t) = h(t)  x(t)
Example 2.1.1 Consider the convolution of the two rectangular pulse signals
t ; 5
x(t) = 2rect 2
t ; 2
h(t) = rect 4
Z1
x(t)  h(t) = x()h(t ; )d
;1
Calculating x(t)  h(t) is conceptually no more di cult than ordinary integration when the
two signals are continuous for all t. Often, however, one or both of the signals is dened in a
piecewise fashion, and the graphical interpretation of convolution becomes especially helpful. We
list in what follows the steps of this graphical aid to computing the convolution integral. These
steps demonstrate how the convolution is computed graphically in the interval ti;1 t ti ,
where the interval ti;1
ti ] is chosen such that the product x()h(t ; ) has the same analytical
form. The steps are repeated as many times as necessary until x(t)  h(t) is computed for all t.
Step 1 For an arbitrary, but xed value of t in the interval ti;1
ti], plot x(), h(t ; ), and the
product x()h(t ; ) as a function of . Note that h(t ; ) is a folded and shifted version
of h() and is equal to h(;) shifted by t seconds.
Step 2 Integrate the product x()h(t ; ) as a function of . Note that the integrand depends on
t and , the later being the variable of integration, which disappears after the integration
is completed and the limits are imposed on the result. The integration can be viewed as
the area under the curve represented by the integrand. The convolution of x(t) and h(t) is
graphically presented in Figure 2.3.

B. Boashash, Signals in Linear Systems


2.2. Properties of the Convolution Integral 17
2.2 Properties of the Convolution Integral
Z1
Given the convolution x(t)  h(t) = x()h(t ; )d we have the following.
;1
 x(t)  h(t) = h(t)  x(t) commutativity
This property is easily proved by variable substitution. It
indicates that the roles of input signal and impulse response
are interchangeable.
x(t) - h(t) - y(t) h(t) - x(t) - y(t)

 x(t)  (h1 (t) + h2 (t)) = x(t)  h1 (t) + x(t)  h2 (t) distributivity


This follows from the linearity property of the integral. It
indicates that a parallel combination of LSI systems is equiv-
alent to a single system whose impulse response is the sum
of individual impulse responses in the parallel conguration.
- h1 (t)
?
x(t) - +l - y(t) x(t) - h1 (t)+h2 (t) - y(t)
6
- h2 (t)

 x(t)  (h1 (t)  h2 (t)) = (x(t)  h1 (t))  h2 (t)


= x(t)  h1 (t)  h2 (t) associativity
This property is derived by changing the order of integration.
It indicates that a cascade combination of LSI systems can
be replaced by a single system whose impulse response is the
convolution of the individual impulse responses.

x(t) - h1 (t) - h2 (t) - y (t) x(t) - h1 (t) h2 (t) - y(t)

Z1
 x(t)  (t) =
;1
x() (t ; ) = x(t)
Z1 Zt
 x(t)  u(t) =
;1
x()u(t ; )d =
;1
x()d integrator]

 x(t)  0 (t) = x0(t), where 0(t) is the derivative of (t)


 u(t)  0 (t) = (t)
 if z (t) = x(t)  y(t), then z 0 (t) = x(t)  y0 (t) = x0 (t)  y(t)
We will see later that these properties are very important in linear systems analysis.
B. Boashash, Signals in Linear Systems
18 Chapter 2. System Analysis in the Time Domain
2.3 Step Response of a Linear System
Sometimes it is very dicult to measure or calculate the impulse response of a linear system.
Using
h(t) = (t)  h(t) = u(t)  h(t)]0
= u0 (t)  h(t) = (t)  h(t) = h(t)
we can simply measure the step response a(t) of the system and calculate its derivative to
determine the impulse response of the system.
y(t) = x(t)  h(t) = x(t)  h(t)  (t)
= x(t)  h(t)  u0 (t) = x(t)  u(t)  h(t)]0
= x(t)  u0 (t)  h(t) = x0 (t)  |u(t) {z
 h(t)]}
step response
= x0 (t)  a(t)
Thus, in terms of the steps response a(t), the response of a system to an input x(t) is the
convolution of the derivative of the input with the step response.

2.4 Stability of Linear Systems


One of the important considerations in any system design is the question of stability.
Denition 2.4.1 A system is bounded-input, bounded-output (BIBO) stable if and only if every
bounded input results in a bounded output.
Theorem 2.4.1 A su cient condition for a xed linear system with impulse response h(t) to
be BIBO is that h(t) satises
Z1
;1
jh(t)jdt < 1

i.e. if h(t) is absolutely integrable.


Proof. The output of a linear time-invariant system can be expressed as
Z1
y (t ) = x()h(t ; )d
;1
then

Z 1

Z 1
jy(t)j =

;1 x()h(t ; )d

;1 jx()jjh(t ; )j d
We require that the input is bounded, thus
jx()j M < 1

where M is nite. Then,


Z1
jy(t)j M ;1
jh(t ; )jd < 1
B. Boashash, Signals in Linear Systems
2.4. Stability of Linear Systems 19
If
Z1
;1
jh(t)jdt = K < 1
then
jy(t)j MK < 1
and the output is then also bounded.
Example 2.4.1 Show that a system with impulse response
1 exp
 t 
h(t) = RC ; RC u(t)
is BIBO stable.
Z1 Z1
;1
jh(t)jdt = ;0
exp(;x)dx = ; exp(;x) j1
0 =0+1=1<1

2.5 Causality of linear systems


theorem 2.4.2: A linear SI system is causal if and only if its impulse
response h(t) is zero for t negative.

B. Boashash, Signals in Linear Systems


20 Chapter 2. System Analysis in the Time Domain

6x()
2

-
4 5 6 
6h()

1
-
4 
6h(;)

1
-
;4 
6h(t ; )

1
-
t;4 t 
Figure 2.2: The plot of x(), h(), h(;), and h(t ; ) as a function of .

B. Boashash, Signals in Linear Systems


2.4. Stability of Linear Systems 21

6 case 1

-
t;4 4 t 6 
6 case 2

-
t;4 4 6 t 
6 case 3

-
4 t;4 6 t 
6 x(t)  h(t) 
 A
A
 A
 A
 A
 A
 A


A
A -
4 6 8 10 t
Figure 2.3: Convolution of x(t) and h(t).

x(t) - h(t) - y(t) = x(t)  h(t)

Figure 2.4: Linear shift-invariant system.

B. Boashash, Signals in Linear Systems


22 Chapter 2. System Analysis in the Time Domain

B. Boashash, Signals in Linear Systems


Chapter 3

Spectral Representation of Signals


3.1 The Fourier Series
3.1.1 Introduction
Orthogonal representations play a crucial role in engineering and science. They are mathemati-
cally convenient to represent arbitrary signals as a simple weighted sum of orthogonal waveforms,
and the calculations involving signals are simplied by using such a representation. The sig-
nal may be represented as a vector in an orthogonal coordinate system with the orthogonal
waveforms being the unit coordinates.
A set of signals i (t), i = 0
1
2
: : : , are said to be orthogonal over an interval a
b] if
Zb 
l (t) k (t) = E

0

k
l=k
l 6= k
a
= Ek l;k
where 
k = 10

kelsewhere
=0

is Kronecker's delta function. If the constant Ek = 1, the i (t) are said to be orthonormal
signals.
Orthogonal signals are useful in that they lead to a series representation of signals in a
relatively simple fashion. Let i (t) be an orthonormal set of signals on an interval a t b
and let x(t) be a given signal with nite energy over the same interval. We can represent x(t)
in terms of f i (t)g by the series
1
X
x(t) = Xi i (t)
i=;1
where
Zb
Xk = x(t) k (t)dt
k = 0
1
2
: : :
a
Below, we investigate a method for expressing periodic signals in terms of \harmonically re-
lated"1 complex exponentials. The choice of complex exponentials as an orthogonal basis is
appropriate since the complex exponentials are periodic, relatively easy to manipulate mathe-
matically, and the result has meaningful physical interpretation.
1 Harmonics will be dened later in the text

23
24 Chapter 3. Spectral Representation of Signals
3.1.2 Trigonometric Fourier Series Representations for Periodic Signals
Let x(t) be a periodic signal with period T0 with f0 = 1=T0 . The general form of the Fourier
series representation of x(t) is the following
x(t) = a0 + a1 cos(2f0 t) + a2 cos(4f0 t) +

+ b1 sin(2f0 t) + b2 sin(4f0 t) +


;1 < t < 1
which can be rewritten as
1
X 1
X
x(t) = a0 + an cos(2nf0 t) + bn sin(2nf0 t)
;1 < t < 1 (3.1)
n=1 n=1

f0 : fundamental
2f0 : 1st harmonic
..
.
nf0 : (n ; 1)th harmonic
The problem is to nd a0 , an , and bn for a given x(t) and T0 . Then we shall obtain the
representation of x(t). Using
Z T0 
sin(m2f0 t) sin(n2f0 t)dt = 0
m 6= n
T0
m=n
Z 0T0  2

cos(m2f0 t) cos(n2f0 t)dt = 0


m 6= n
T0
m=n
0 2

and
Z T0
sin(m2f0 t) cos(n2f0 t)dt = 0
for all m
n 2 N
0

we easily nd that


1 Z T0
a0 = T x(t)dt (3.2)
0 0

and for m 6= 0
2 Z T0
am = T x(t) cos(m2f0 t)dt (3.3)
Z T0
0 0

b = 2
m T0 x(t) sin(m2f t)dt 0 (3.4)
0

This Fourier series will converge to the function itself if it satises the Dirichlet conditions.
Dirichlet conditions: If
(i) x(t) is bounded, absolutely integrable
Z h+T0
h
jx(t)jdt < 1
and of period T0 and
B. Boashash, Signals in Linear Systems
3.1. The Fourier Series 25
(ii) x(t) has at most a nite number of maxima and minima in one period and a nite number
of discontinuities,
then the Fourier series of x(t) converges to x(t) at all points where x(t) is continuous, and
converges to the average of the right-hand and left hand limits of x(t) at each point where x(t)
is discontinuous, i.e.

x(t0 ) = 21 x(t+0) + x(t;0 )]

where t0 is the time where x(t) is discontinuous. Note that the conditions (i) and (ii) are
sucient but not necessary.

3.1.3 The Exponential Fourier Series


Let x(t) be a periodic signal with fundamental period T0 , i.e.
x(t) = x(t + T0 )
for all t 2 R
The use of !0 = 2=T0 and f0 = 1=T0 referred to as the fundamental radian frequency and
fundamental frequency, respectively. Consider the representation of x(t) by an orthogonal set
of basis functions, n (t) = exp(j 2nt=T0 ). Then we can represent x(t) as
1
X  2nt 
x(t) = Xn exp j T (3.5)
n=;1 0

where
1 ZT  2nt 
Xn = T x(t) exp ;j T dt (3.6)
0 0

are complex constants. Each term of the series has a period T0 and a fundamental radian
frequency !0 . Hence, if the series converges, its sum is periodic with period T0 .2 Such a series
is called the complex exponential Fourier series and the Xn are called the Fourier coecients.

3.1.4 Properties of the Fourier Series Coecients


Symmetry
For real-valued signals,
1 Z T  2nt  
Xn = T 0 x(t) exp ;j T dt
1 ZT  2(;n)t 
= T x(t) exp ;j T dt
0
= X;n
2 If x1 (t) and x1 (t) are two periodic signals with period T0 , then x3 (t) = ax1 (t) + bx2 (t) is also periodic with
period T0 .

B. Boashash, Signals in Linear Systems


26 Chapter 3. Spectral Representation of Signals
Hence jX;n j = jXn j and argX;n = - argXn . For a real-valued signal one can rewrite
;1
X  2mt  X 1  2mt 
x(t) = X0 + Xm exp j T + Xm exp j T
m=;1 m=1
X1  2nt   2nt 
= X0 + <fXn g cos T ; 2=fXn g sin T
n=1
X1  2nt   2nt 
= a0 + an cos T + bn sin T ]

n=1
where
ZT
a0 = X0 = T1 x(t)dt
0
2 ZT  2nt 
an = 2<fXn g = T x(t) cos T dt
2
0
ZT  2nt 
bn = ;2=fXn g = T x(t) sin T dt
0

In terms of magnitude and phase of Xn , the real-valued signal x(t) can be expressed as
X1  
x(t) = X0 + jXn j cos 2nt T + argXn (3.7)
n=1
X1  2nt 
= X0 + An cos T + "n (3.8)
n=1
where
An = 2jXn j = 2jX;n j
and
"n = argXn
This representation is alternative to the previous one and is more compact and meaningful.
Thus, for real signals, the magnitude of the Fourier coecients is an even function of the index
n. (where n represents the frequency f)
If x(t) is real and even, i.e. x(t) = x(;t), the imaginary part of Xn is zero, because
x(t) sin(2f0 t) is an odd function that integrates over an interval symmetrically placed about
t = 0. Therefore:
 If x(t) is real and even, Xn is real and even. (ie Im[Xn]=0)
 Similarly, if x(t) is real and odd, Xn is imaginary and odd. (ie Re[Xn]=0)
Example 3.1.1 Consider the following signal
8
< 0
;T=2 < t < ; =2
x(t) = : K
; =2 < t < =2 x(t) = x(t + T )
0
=2 < t < T=2
B. Boashash, Signals in Linear Systems
3.1. The Fourier Series 27
Signals of this type can be produced by a pulse generator and are used extensively in radar and
sonar systems. Calculate the Fourier coe cients of x(t).
1 Z T=2  2nt 
Xn = T x(t) exp ;j T dt
;T=2
 
1 Z =2 K exp ;j 2nt dt
= T ;=2 T
  
K T j exp ;j 2n ; exp +j 2n

= T 2n 2T 2T

j K 2j sin 2n
= ; 2n 2T
= K sin  n  = K T sin  n 
n T n T T
= K sinc  n 
T T
where
sinc(x) = sin(x) :
x
Parseval's Theorem
The averaged normalised power of a periodic waveform x(t) is given by
1 Z 1 Z
Pav = T
0 <T0 >
jx(t)j dt = T
2
0 <T0 >
x(t)x (t)dt

R
where <T0 > denotes integration over interval of length T0 . If we replace x(t) by its Fourier
series representation, we get
Z  1 !
1 X
Pav = T x(t) N X  e;j 2nf0 t
dt
0 <T0 >
n=;1
X1 1 Z
= 
Xn T x(t)e ; j 2nf0 t
dt
n=;1 0 <T0 >

X1
= Xn Xn
n=;1
1
X
= jXnj2

n=;1
so that
Z 1
X
Pav = T1 jx(t)j dt =
2
jXnj :
2
0 <T0 > n=;1
The average power of a periodic signal x(t) is the sum of the powers in the components of its
Fourier series.
B. Boashash, Signals in Linear Systems
28 Chapter 3. Spectral Representation of Signals
Periodic Convolution
Let x(t) and y(t) be two periodic signals with same period fundamental period T0 . We dene
the circular convolution of x(t) and y(t) by
1 Z
z (t ) = T x( )y(t ; )d
(3.9)
0 <T0 >
where the signal is taken over one period T0 . One can show that
1 Z
z (t + T0) = T x( + T0 )y(t ; + T0 )d
0 <T0 >
1 Z
= T0 x( )y(t ; )d
<T0 >
= z (t)
and therefore z (t) is periodic with period T0 . The reader should show that the periodic convo-
lution is commutative and associative. We can write z (t) in a Fourier-series representation with
coecients
1 Z T0
Z =n T0 z(t)e;jn2f0 tdt
1 Z0
T0 1 Z T0
T0 0 x( )y(t ; )e
= T ;j 2nf0 t d dt
Z T0
0 0
Z T0
= T1 x( )e;j2nf0  T1 y(t ; )e;j2f0 (t; ) dt d
1
0 0
Z T0 10 Z0 T0 ;
= T x( )e ; j 2nf0 
T0 ; y(v)e
;j 2f0 v
(;dv) d
0 0
Z T0
= T1 x( )e;j2nf0  Ynd = Xn Yn
0 0

where Xn and Yn are the Fourier series coecients of x(t) and y(t), respectively.
3.1.5 Representation of the Fourier Coecients in the Frequency Domain
Line Spectra. We have seen that if x(t) is periodic with period T0, then
1
X
x(t) = Xn ej 2nf0t
n=;1
Z
Xn = T1 x(t)E ;j2nf0 t
0 <T0 >
and we have (for a real signal)
Xn = X; n
jXnj = jX;nj
"n = ;";n:
This observation allows a periodic signal to be characterised graphically in the frequency domain
by two plots
B. Boashash, Signals in Linear Systems
3.1. The Fourier Series 29
 One showing amplitudes of the components versus frequency, which is known as the am-
plitude spectrum of the signal.
 The other showing the relative phase of each component versus frequency, which is called
the phase spectrum of the signal.
 We see that these spectral components or lines, are present at both positive and negative
frequencies.
 For a real signal, the amplitude spectrum is even and the phase spectrum is odd.
Example 3.1.2 (Pulse-train signal.) Let
1
X  t ; nT 
x(t) = A rect
0
:
n=;1
x(t)
6

-
;T 0 ; 2

2 T0 t
Since x(t) is periodic with period T0 it may be represented as a Fourier series, with fundamental
period T0 and fundamental frequency f0
1
X
x(t) = Xn ej 2nf0t :
n=;1
Solving for the Fourier series coe cients, we found in example 3.1.1
A
 n 
Xn = T sinc T :
0 0

1. We note that the width of the envelope of the amplitude spectrum increases as the pulse
width decreases (for xed T0 ). That is, the pulse width of the signal and its corresponding
spectral width are inversely proportional.
2. We also note that the separation between lines in the spectrum is T10 , and therefore the
density of the spectral lines with frequency increases as the period T0 of x(t) increases (for
xed ).
Example 3.1.3 We which to compute and plot the magnitude and phase of the Fourier series
coe cients of the signal given in Figure 3.1.

x(t) = ;k
;1 < t < 0
k
0<t<1
We have x(t + 2) = x(t) and therefore !0 = 2=2 = .
8
< 2k
n odd
Xn = : jn
0
n even
B. Boashash, Signals in Linear Systems
30 Chapter 3. Spectral Representation of Signals
x(t)
6
k
-
0 1 t

Figure 3.1: Periodic signal x(t)

which leads to
8 2k
<
jXnj = : jnj
n odd
0
n even
and
8 
< ;2
n = 2m ; 1
m = 1
2
: : :
argXn = : 0
n = 2m
m = 0
1
2
: : :


2 n = ;(2m ; 1)
m = 1
2
: : :

s j j
6Xn s 2k

s s 2k
3
s s 2k
5
s s s s s -
0 1 2 3 4 5 n

j j
6arg Xn
s s s 
2

s s s 1 2s 3 4s 5 -
0 n

;
s
2
s s

3.1.6 Systems with Periodic Inputs


Consider a linear time-invariant system with impulse response h(t). If x(t) is the system's
driving signal then its output is
Z1
y (t ) = h( )x(t ; )d :
;1
B. Boashash, Signals in Linear Systems
3.1. The Fourier Series 31
Let x(t) = ej 2ft . The output of the system is then given by
(Case 1:) Z1
y(t) = h( )ej2f (t; ) d
;1 Z
1
= e 2ft
j h( )e;j2f d :
;1
We can then write
y(t) = H (f )ej2ft (3.10)
where H (f ) is the system transfer or frequency response function and is a constant for xed f ,
Z1
H (f ) = h(t)e;j2ft dt:
;1
Rewriting H (f ) = jH (f )jej
H (f ) , where jH (f )j is called the amplitude-response and "H (f ) the
phase response function, respectivly, we nd
y(t) = jH (f )jej(2ft+
H (f )) :
This results shows that the frequency response H (f ) completely characterises the response of
a time-invarint linear system to an exponential ej 2ft , which is called eigenfunction of the LTI
system. H (f ) is the eigenvalue of the system.
Case 2: General
To determine the response y(t) of an LTI system to a periodic input x(t), we use the Fourier
series representation and the previous result.
If x(t) has period T0 , it can be written as
1
X
x(t) = Xn ej 2nf0 t
f0 = T1
n=;1 0

which leads together with (3.10), and using the linearity property, to
1
X
y(t) = H (nf0) Xn ej2nf0 t :
n=;1
This equation indicates that the output signal is the summation of exponentials with coecients
H (nf0) Xn
Note that since H (nf0 ) is a complex constant for each n, it follows that the output is also
periodic with Fourier-series coecients H (nf0 ) Xn . Since the fundamental frequency of x(t) is
T0 = f0 , the period of y(t) is equal to the period of x(t). Hence, the response of an LTI system
1

to a periodic input with period T0 is periodic with the same period.


The output can be re-written in real form as
1
X
y (t ) = jH (f )jan cos(2nf t + "H (nf )) + bn sin(2nf t + "H (nf ))]
0 0 0 0 0
n=0

B. Boashash, Signals in Linear Systems


32 Chapter 3. Spectral Representation of Signals
3.1.7 Summary
 Jean Baptiste de Fourier showed that any periodic signal could be represented through an
innite sum of sines and cosines, with the fundamental period of such functions equalling
that of the signal itself. This ensures that the resulting Fourier series must have the same
period as the signal.
 As a result of the orthogonality property of sines and cosines, the Fourier series coecients
may be conveniently calculated.
 The Fourier series may be reformulated in terms of the complex exponential ej f0t = 2

cos 2f0 t + j sin 2f0 t:


 The Fourier series provides a frequency domain characterisation of a periodic signal. This
is represented in terms of an amplitude function and a phase function.

3.2 The Fourier Transform (for any signal)

3.2.1 De
nition of the Fourier Transform
We have seen in the previous section that the Fourier Series representation of a periodic signal
x~(t) is
1
X
x~(t) = Xn ej 2nf0 t (3.11)
n=;1
where
1 Z T0 =2
Xn = T x~(t)e;j 2nf0t dt (3.12)
0 ;T0 =2
To develop a spectral representation for non-periodic signals (i.e. signals with period T0 !
1), we let the frequency spacing f0 = T0;1 approach zero such that f0 ! df , an innitesimally
small quantity, and the product nf0 approaches a continuous frequency variable f .
6 6
x~(t)

T0 -
x(t)

- -
t t

Figure 3.2: Allowing period T0 to increase to obtain the non-periodic signal.


Let x(t) be the signal x~(t) when T0 ! 1. Then using (3.12) we can write
Xn = Z 1 x~(t)e;j 2ft dt (3.13)
df ;1

B. Boashash, Signals in Linear Systems


3.2. The Fourier Transform 33
Substituting (3.13) into (3.11), and recognising that in the limit the sum becomes an integral
and x~(t) approaches x(t) we obtain
Z 1 Z 1
x(t) = x~(t)e;j 2ft dt ej 2ft df: (3.14)
;1 ;1
The inner integral is a function of f only, not t. Denoting this integral by X (f ), the previous
equation reduces to
Z1
x(t) = X (f )ej2ft df (3.15)
;1
where
Z1
X (f ) = x(t)e;j 2ft dt: (3.16)
;1
These expressions dene a Fourier Transform pair. It is sometimes denoted:

X (f ) = F fx(t)g
x(t) = F ;1 fX (f )g
or x(t) $ X (f )
3.2.2 Convergence of the Fourier Transform
Sucient conditions for the convergence of the Fourier Transform integral are that:
Z1
1.
;1
jx(t)jdt exists, indicating that signals are absolutely integrable. This property comes
from the fact that the Fourier transform must be bounded, i.e.

x(t)e; j 2 ft dt

jx(t)jdt < 1:

2. Any discontinuities in x(t) be nite, and the number of such discontinuities must be nite.
This property is referred to as x(t) is \well behaving". Except for impulses, most signals
of interest are well behaved and satisfy (3.15).
These conditions exclude a number of important signals such as sinusoids that are not abso-
lutely integrable. By means of distributions (delta functions), signals of this type can be handled
using essentially similar methods as for nite energy signals.
Example 3.2.1 The Fourier transform of the rectangular pulse x(t) = rect(t= ) is
Z1
X (f ) = x(t)e;j2ft dt
;1
Z =2
= e;j2ft dt
;=2
j e;jf ; ejf 
= 2f
= sinc(f )

B. Boashash, Signals in Linear Systems


34 Chapter 3. Spectral Representation of Signals
Example 3.2.2 Consider the triangular pulse dened as
8
< jtj
$(t= ) = : 1 ;
jtj
0
jtj >
This pulse is of unit height, centered about t = 0, and of width 2 . The Fourier transform is
Z1
X (f ) = $(t= )e;j 2ft dt
;1 
Z0 t Z   t
= 1+ e ;j 2ft
dt + 1 ; e;j 2ft dt
Z;  t  Z 0 t 
= 1 ; ej 2ft dt + 1 ; e;j 2ft dt
Z   t
0 0

= 2 1 ; cos(2ft)dt
0
= sinc2(f )
Example 3.2.3 The Fourier transform of the impulse function is
Z1
Ff (t)g = ;1
e;j 2ftdt = 1

Using the inversion formula we have


Z1
(t ) = ej 2ft df
;1 Z

= lim
!1 ej2ft df
;
= lim 2 sin(t)
!1 t
3.2.3 Fourier Transforms in the Limit
In the ordinary sense, there are signals which do not possess a Fourier transform. For example,
a sinusoid, x(t) = sin(2ft), extending over all time is not absolutely integrable. Another
example is the impulse function which has innite discontinuities. This motivates broadening
the denition of the Fourier transform, to include Fourier transforms in the limit.
Consider x(t) = A, where A is a constant. It is clear that its Fourier transform does not
exist. But the Fourier transform of Arect(t=T ) does exist and is

FfArect(t=T )g = A sinffT = AT sincfT


If T ! 1, we'll obtain an oscillatory function withR main lobe cantered at f = 0 which becomes
very narrow and high for large T . Furthermore, ;1 1 AT sinc(fT )df = A for any T . Thus for
T ! 1, we formally write that FfAg = A (f ), since limT !1 AT sinc(fT ) = A (f ) (This is
another way to approach a delta function in the limit).
B. Boashash, Signals in Linear Systems
3.2. The Fourier Transform 35
Example 3.2.4 Consider the exponential signal x(t) = ej2f0 t. The Fourier transform of this
signal is

Z1
X (f ) = ej2f0t e;j 2ftdt
Z;1
1
= e;j 2(f ;f0 )t dt
;1
= ( f ; f0 )

This is an expected result since ej 2f0 t has energy concentrated at f0 .

Table 3.1: Some selected Fourier transform pairs.

x(t) X (f )
1. 1 (f )
2. u(t) 0:5 (f ) + 1=(j 2f )
3. (t) 1
4. (t ; t0 ) e;j2f0 t
5. rect(t= ) sin(f )=(f )
6. sin(2fB t)=(t) rect(f=2fB )
7. ej 2f0 t (f ; f0)
1
X X1
8. anejn2f0 t an (f ; f0)
n=;1 n=;1
9. cos(2f0 t) 0:5 (f ; f0 ) + (f + f0 )]
10. sin(2f0 t) ;j0:5 (f ; f ) ; (f + f )]
0 0

11. e;at u(t), <fag > 0 1=(a + j 2f )


12. te;at u(t), <fag > 0 1=(a + j 2f )2
13. e;ajtj , a > 0 2a=(a2 + 42 f 2 )
14. $(t= ) sinc2 (f )

B. Boashash, Signals in Linear Systems


36 Chapter 3. Spectral Representation of Signals
3.2.4 Energy Spectral Density
The energy of a signal can be expressed in the frequency domain. The normalized energy for a
signal is
Z1
E =
;1
jx(t)j 2
dt
Z1
E = x(t)x(t)dt
Z;1
1 Z 1
E = 
x (t) j 2ft
X (f )e df dt
Z;1
1 Z ;1
1
E = X (f ) x(t)ej 2ft dt df
Z;1
1 Z;11 
E = X (f ) x(t)e ; j 2ft
dt df
;1
Z1 ;1 Z1
E =
;1
X (f ) X  (f ) df = jX (f )j2 df
;1
R1
E = ;1 jx(t)j2 dt = R;1
1 jX (f )j2 df

This is Parseval's theorem for Fourier transforms. jX (f )j2 is the energy spectral density of
x(t). Integration over all frequencies yields the total energy contained in a signal. Integration
over a nite range B of frequencies gives the part of energy contained in this range B . Likewise
for jx(t)j2 {known as the instantaneous power.
Example 3.2.5 Consider the one-sided exponential signal
x(t) = e;t u(t)
From Table 3.1 we have
jX (f )j 2
= 1 + 412 f 2 :
The total energy in this signal is
Z1
E =
;1
jX (f )j df
2

Z1 1 df
=
;1 1 + 42 f 2
= 12

3.2.5 Properties of the Fourier Transform


 Linearity (Superposition) Theorem: obvious because it is an integral operation.
 Time Delay Theorem: x(t ; to) $ X (f )e;j ft 2 o

B. Boashash, Signals in Linear Systems


3.2. The Fourier Transform 37
 
 Scale change: x(at) $ jaj; X fa 1 (time-scale methods & wavelets)

 Time reversal: x(;t) $ X (;f ) = X (f )


 Duality: X (t) $ x(;f )
 Frequency translation: x(t)ej F t $ X (f ; fo)2 o

 Modulation: x(t) cos !ot $ X (f ; fo) + X (f + fo)


1
2
1
2

 Dierentiation: d dtx t $ (j2f )nX (f )


n ()
n

 Integration: R;1
t x(t0 )dt0 $ (j 2f ); X (f ) + X (0) (t) 1 1
2

 Convolution: x (t)  x (t) $ X (f )


X (f )
1 2 1 2

 Multiplication: x (t)
x (t) $ X (f )  X (f )
1 2 1 2

 Proof of the convolution theorem:


Z1
x1 (t)  x2 (t) = x1 (t ; )x2 ()d
;1

now
Z1
x1 (t ; )e;j 2ft dt = X1 (f )e;j 2f
;1
therefore
Z1
x1(t ; ) = X1 (f )e;j2f ej2ft dt
;1

Z1Z1
) x (t)  x (t) =
1 2
;1 ;1
X1 (f )x2 ()e;j2f dej 2ft dt

Z1
= X1 (f )X2 (f )ej2ft dt = F ;1 X1 (f )X2(f )]
;1

therefore
F x (t)  x (t)] = X (f )X (f )
1 2 1 2

B. Boashash, Signals in Linear Systems


38 Chapter 3. Spectral Representation of Signals
Fourier Transform of Special Functions:
 we know that 1 $ (f )
 duality theorem: (t) $ 1(;f ) = 1
 (t ; to) $ exp(;j2fto)
 exp(j2fot) $ (f ; fo)
 cos 2fot $ (f ; fo) + (f + fo)
1
2
1
2

 sin 2fot $ j ( (f ; fo) ; (f + fo))


1
2

 u(t) $ 1
(j2f ); + (f ) because u(0) = 1 and u(t) = R;1
1 1
2
t (t)dt (integration theorem).

 sgn (t) $ (jf ); (use sgn (t) = 1 ; 2u(;t))


1

 t $ ;j sgn f (duality)
1

Example 3.2.6
1
X 1
X
y(t) = (t ; mTs ) $ fs (f ; mfs )  fs = Ts;1
m=;1 m=;1

Proof: this signal is a periodic train of impulses called the ideal sampling waveform. Its
Fourier series is

1
X
y(t) = Yn ej2nfs t
n=;1
R
with Yn = T1s ;TTs =s2=2 (t) e;j 2nfs t dt = fs
P
therefore, y(t) = 1 ;1 fs e
j 2nfs t

1
X 1
X
y(t) = fs ej 2nfs t $ fs (f ; fs)
;1 ;1
Example 3.2.7 Application of the dierentiation theorem.
Consider the triangle signal:

$(t=T ) = 1 ; jTtj
jtj < T
0
elswhere

d ; T sgn(t)
1

jtj < T
dt $(t=T ) = 0
elsewhere
B. Boashash, Signals in Linear Systems
3.3. The Laplace Transform 39

d $(t=T ) = 1 rect (t + T ) ; rect (t + T ) = ; 1 rect (t) sgn(t)


dt T T 2 T 2 T 2T

d2 $(t=T ) = 1 (t + T ) ; 2 (t) + 1 (t ; T )
dt T T T
then

d2 $(t=T ) $ (j 2f )2
F $(t=T )] = 1 hej2fT ; 2 + e;j 2ft i
dt T

;! $(t=T ) $ (j2f1 ) T ;2 + 2 cos 2fT ] = T sin(fTfT)


2
2 2

where sin " = (1 ; cos 2") has been used.


2 1
2

;! $(t=T ) $ T sinc fT 2

3.3 The Laplace Transform: Causality, Stability and Physical


Realisability
The Fourier transform of unstable signals does not exist. For such signals, the Laplace trans-
form may be employed. The Laplace transform is essentially a generalised form of the Fourier
transform, utilising a broader class of complex exponential signals, e;st , where s =  + j 2f .
The Laplace transform may converge for some values of  but not for others. If it converges for
 = 0 then the Fourier transform also exists for such a signal.
3.3.1 Existence of the Laplace transform
Sucient condition to assure the convergence of integral are
Theorem 3.3.1 If x(t) is integrable in every nite interval 0 a < t < b < 1 and for some
value 0 the limit
lim e; 0 t jx(t)j
t!1
exists, then the Laplace integral converges absolutely and uniformly for <fsg > 0 , and is dened
by
Z1
X (s ) = x(t)e;st dt < 1:
0

We shall also use the operator notation X (s) = Lfx(t)g and sometimes x(t) $ X (s).
When a Laplace transform is computed, there is a restriction on the value of  for which
the transform is valid. This value determines the region in the complex s-plane in which the
integral converges.
B. Boashash, Signals in Linear Systems
40 Chapter 3. Spectral Representation of Signals

6
1 ^T
;;@@
; @
; @ -t
;T T
d ^
dt T
1 6
T

-t

;1
T d2^
dt2 T
6
1 1
T T
6 6 -t

;2 ?
T

Figure 3.3: Triangular signal and its derivatives

6j 2f
The complex s-plane
-

The value of  is important in determining a path of integration that could be used for evaluating
the inverse Laplace transform
I 0+j1
x(t) = j 21 X (s)est ds
0 ;j 1
by contour integration in the complex plane. If 0 = 0, it then follows that the Fourier transform
of x(t) exists and can be obtained form X (s) by substituting s = j 2f .

B. Boashash, Signals in Linear Systems


3.3. The Laplace Transform 41
Examples of Laplace transforms. One can easily show that
Lf (t)g = 1
Lfu(t)g = 1s
Example 3.3.1 Compute the Laplace transform of x(t) = e;t u(t),  2 R. By denition,
L e;tu(t) = s +1 
<fsg > ;
if  > 0, the region of convergence (ROC) is as depicted in Figure 3.4. Note that the Fourier

j 2 f

Figure 3.4: Region of convergence <fsg > ;.


transform of x(t) = u(t)e;t is given by
Ffu(t)e;t g = X (s)js =j 2f =  + 1j 2f

3.3.2 Relation Between Fourier Transform and Laplace Transform


There are many signals of interest for which Fourier transforms do not exist. In this case, we
can dene a new transform as follows
Ffx(t)e; t u(t)g
where x(t) = 0, for t < 0 and   0. This yields
Z1
X ( + j 2f ) = x(t)e; t e;j2ft dt
0

With s =  + j!, it can be written


Z1
Lfx(t)g = X (s) = 0
x(t)e;st dt
 > 0
where L is the Laplace transform operator (single-sided).

B. Boashash, Signals in Linear Systems


42 Chapter 3. Spectral Representation of Signals
Inverse Laplace Transform
1 Z +j1
x(t) = 2j X (s) est ds
;j1
R +j1 st
The integral x(t) = 2j 1
;j1 X (s) e ds is carried out along any line to the right of the singu-
larities of X (s).
If this line can be chosen as the j 2f {axis, it then follows that the ordinary Fourier Transform
of x(t) exists and can be obtained from X (s) by substituting s = j 2f .
The Laplace transform may be used in design or analysis to determine if a system is stable or
to establish valid limitations on the form of a system transfer function. (The Fourier transform
may also be used). The following relationships between the Fourier transform and the Laplace
transform exist.
Relationship between the Fourier transform and Laplace transform
1. If 0 > 0, the axis j 2f is outside the region of convergence ! the Fourier transform does
not exist.
2. If o < 0, there are 2 possibilities, being the existence or non-existence of poles on the
imaginary axis:
(a) if there are no poles ) X (s) ! X (f )
s ! j 2f
(b) if there are N poles sk = j!k = j 2fk where Ak is the residue
X
N
X (f ) = X (s)js=j2f + 21 Ak (f ; fk )
k=1
at the pole sk .

Examples:
Lfu(t)g = 1s
<fsg > 0
Ffu(t)g = j 21f + 12 (f )
Ffej f tu(t)g
2 o
= j 2(f1; f ) + 12 (f ; f0 )
0

Lfej f t u(t)g
2 o
1
= s ; j 2f
0

3.3.3 Conditions for the Realisability of a Linear System or Filter


For a linear system or a lter to be realisable three factors are to be fullled. Namely a realisable
linear system must be causal, real, ans stable.
B. Boashash, Signals in Linear Systems
3.3. The Laplace Transform 43
1. A system is called casual if
h(t) = 0 for t < 0:
In this case the transfer function of a lter is given by
Z1
H (s) = h(t)e;st dt:
0

2. A linear system with impulse response h(t) is real if


h(t) = h(t) for all t 2 R
In the s or frequency domain this property translates to
H (s ) = H  (s):
3. A linear system with impulse response h(t) is stable if
Z1
(a)
;1
jh(t)jdt < 1
(b) all the poles sk of H (s) are such that <fsk g < 0,
(c) if H (s) can be represented as a ratio of two polynomials

H (s) = N (s)
D(s)
then the order of N (s) must be smaller than the order of D(s).
4. jH (s)j is nite for all values of f .
5. jH (s)j ! 0 as f ! 1 and jh(t)j has the same behaviour as f1n , n  1.

Various tests (Routh, Hurwitz) exist to determine if this is so.


3.3.4 Frequency Response of Transfer Function using the Poles and Zeros of
the Laplace Transform
The approximate frequency response of a system can be determined by using the poles and zeros
of the Laplace transform. The frequency response is derived from the Laplace transform as one
moves along the j 2f axis. Along this axis, points close to poles will have large magnitude
responses, and vice versa for all points on the axis near zeros.
Example 3.3.2
H (s) = ss2 ++42ss++13
2 = (s + 1)2 + 1
2

(s + 2)2 + 9

Poles and zeros may be used for preliminary design.


B. Boashash, Signals in Linear Systems
44 Chapter 3. Spectral Representation of Signals
j2 f H(f)/H(0)

-2 -1 2 f
-3 -2 -1 1 2 3

Figure 3.5: System of second order.

3.3.5 Poles and Zeros of the Laplace Transform and its Region of Conver-
gence
In many applications, the Laplace transform (or some part of it) can often be expressed in terms
of a rational polynomial function of the form
Y
N
(s ; zk )
X (s) = N (s) k=1
D(s) = Y
M
M N
(s ; pk )
k=1
The roots of N (s) are the zeros of X (s), since they represent the points in the Laplace transform
plane where X (s) = 0. Similarly, the roots of D(s) are the poles of X (s) since X (s) ! 1 at
such values of s. The poles and zeros of the Laplace transform completely determine the ROC.
Example 3.3.3
x(t) = (t) ; 34 e;t u(t) + 13 e2t u(t)

; 1)2

X (s) = 1 ; 3(s 4+ 1) + 3(s 1; 2) = (s +(s1)( <fsg > 2


s ; 2)
The region of convergence is determined by the poles of the system.
Example 3.3.4 Impulse response from the transfer function.
H (s) = s25+s +6s13+ 5
The roots of the denominator s2 + 6s + 5 are given by s1 = ;1 and s2 = ;5. Therefore

H (s) = (s +5s1)(
+ 13 = A + B
s + 5) s + 1 s + 5
B. Boashash, Signals in Linear Systems
3.3. The Laplace Transform 45
j2 f

ROC

1 1 2

Figure 3.6: System of second order.

To nd A and B we use the residue theorem, and evaluate respectively


A = H (s)
(s + 1)js=;1 = 13 4; 5 = 2
B = H (s)
(s + 5)js=;5 = ;25;+4 13 = 3
which leads to
H (s) = s +2 1 + s +3 5
The inverse Laplace transform of H (s) is then given by
h(t) = 2 e;t u(t) + 3e; 5t u(t)
= 2e;t + 3e;5t u(t):
Example 3.3.5 Consider the following transfer function
H (s) = s +s83s++7s23 s2 + 34s + 19
4 3
2 + 15s + 9

= s + 1 + s3 s+ 7+s210+s 15
+ 10
2

s+9
By inspection, we can nd the rst root of the denominator as s1 = ;1. To nd the remaining
roots, we perform the following division.
s3 + 7s2 + 15s + 9 = s2 + 6s + 9 = (s + 3)2
s+1
Thus
s3 + 7s2 + 15s + 9 = (s + 1)(s + 3)2
and
A + B + C
H~ (s) = s + 1 s + 3 (s + 3)2
B. Boashash, Signals in Linear Systems
46 Chapter 3. Spectral Representation of Signals
A and C is determined as before
A = H~ (s)
(s + 1)js=;1 = 1 ; 104 + 10 = 41
C = H~ (s)
(s + 3)2 js=;3 = 9 ; 30
;2
+ 10 = 11
2
For determining B , we take s = 0, which leads to
1 + B + 11 1 = 10 :
4 3 29 9
Thus B = 34 . Finally,

H (s) = s + 1 + 4(s 1+ 1) + 4(s 3+ 3) + 2(s 11


+ 3)2
Its inverse Laplace transform is then

h(t) = d dt(t) + (t) + 41 e;t u(t) + 34 e;3t u(t) + 11 ;3t


2 te u(t):
Important Results
1. The region of convergence of X (s) consists of strips parallel to the j 2f axis.
2. If x(t) is of nite duration, then the region of convergence is the entire s-plane.
3. If the line <fS g = 0 is in the region of convergence, then all values of s where <fS g > 0
will also be (given by the rightmost pole).

3.4 Time{Varying Spectra (or time-frequency methods)

3.4.1 The Hilbert Transform


Denition 3.4.1 The Hilbert Transform Hfx(t)g of a signal x(t) is obtained by convolving x(t)
with
1
t .

1 1 Z 1 x( )
Hfx(t)g = x(t)  t =  ;1 t ; d (in principle value)

x(t) - h(t) - Hfx(t)g


The lter with impulse response h(t) is non-causal. Using the properties of the Fourier
Transform, it follows that
1
FfHfx(t)gg = F t
Ffx(t)g
B. Boashash, Signals in Linear Systems
3.4. Time{Varying Spectra 47
Knowing that
Ffsgn(t)g = jf1
and because FfY (t)g = y(;f ), (the duality principle), we have
1
F = sgn(;f ) = ;sgn(f )
jt
Therefore,
1
F t = ;j sgn(f )
3.4.2 Analytic signals
Denition 3.4.2 An anlaytic signal is a complex{valued signal whose spectrum is zero for neg-
ative frequencies. It contains only positive frequencies.
Theorem 3.4.1 A signal z(t) is analytic if and only if it is complex with its real and imaginary
parts forming a Hilbert transfrom pair.
Proof 3.4.1 Let us consider the signal z (t) for which Z (f ) = 0, for f < 0.
Step 1: z (t) must be complex, otherwise if z (t) was real, its FT would have Hermitian property and
contradict the hypothesis.
Step 2:
Let z (t) = x(t) + j:y(t): Its FT is :
Then Z (f ) = X (f ) + j:Y (f ) = 0
for f < 0
;! Y (f ) = j:X (f ) for f < 0 :
(3 17)
;! Y (;f ) = j:X (;f ) due to Hermitian property of FT of real signals
;! Y (f ) = j:X (f ) for f > 0
;! Y (f ) = ;j:X (f ) for f > 0 :
(3 18)
Equations (3.17) and (3.18) can be combined in one as:
Y (f ) = ;j sgn f:X (f )
As the FT of (;j sgn f ) is 1
t , this becomes:
y(t) = Hx(t)] = t1 x(t) QED
Note also that if f > 0
jY (f )j = jY (;f )j = jjX (;f )j = jX(;f )j   = jX (f )j 
"Y (f ) = ;"Y (;f ) = ;"jX (;f ) = ; "X (;f ) + 2 = "X (f ) ; 2
Then, f > 0, Y (f ) = ;jX (f ), so that Y (f ) = ; j sgn f X (f ), for all f and Z (f ) =
2U (f ) X (f ), for all f where U (f ) = 1, for f > 0 and 0 for f < 0.
Analytic signals are of use in modulation theory applications.
Application. One application of analytic signals is in Single-Sideband AM (SSB-AM). Consider
the spectrum of an unmodulated real information signal, s(t), as shown in Figure 3.7. Since
the signal s(t) is real, the information in the left and right hand sides of the spectrum is the
same (symmetrical). For transmission, it must be modulated by a carrier cos(2f0 t), to give
r(t) = s(t) cos(2f0 t). Its spectrum R(f ) is shown in Figure 3.8. If instead we take the analytic
signal of s(t), z (t) = s(t)+ j Hfs(t)g i.e. only the positive frequencies, we obtain after modulation
the spectrum shown in Figure 3.9.
B. Boashash, Signals in Linear Systems
48 Chapter 3. Spectral Representation of Signals
S(f)

0 f

Figure 3.7: Spectrum of real information signal.


R(f)

f0 0 f0 f

Figure 3.8: Spectrum of modulated information signal.

3.4.3 Time{Frequency Relationships


Let x(t) be a deterministic signal and X (f ) its Fourier Transform.
 The dimension of the variable t is TIME
 The dimension of the variable f is (TIME); 1
= 1= TIME
Each time we use the Fourier Transform, we obtain these two kinds of parameters, with the
conseqence that we cannot localize simultaneously x(t) and X (f ).

Assumptions on the Signals


 x(t) is centered around t = 0
 X (f ) is centered around f ; 0
 x(t) 2 L , X (f ) 2 L
2 2

Let p(t) be the energy density versus time t:

p(t) = jxk(xt)kj 0
2
real

p(t) has the dimensions of a probabilty density. The \time interval where it is concentrated"
is given by $t such that equivalent duration:

1 Z1
($t) = kxk
2
t2 jx(t)j2 dt
;1
Similarly in the frequency domain with $y equivalent bandwidth.
B. Boashash, Signals in Linear Systems
3.4. Time{Varying Spectra 49
R(f)

f0 0 f0 f

Figure 3.9: Spectrum of real information signal.

Figure 3.10:

Time-frequency inequality
R1 2
t jx(t)j2 dt
($t) = ;1
2
R1 equivalent duration
;1 jx(t)j dt
2

R1 2
f jx(f )j2 df
($f ) = ;1
2
R1 equivalent bandwidth
;1 jx(f )j df
2

kxk = kzk (Parseval's theorem)


dx(t) = x_ (t) *
) 2jfX (f )
dt

Z1 Z1
! ;1
jx_ (t)j2
dt =
;1
42 f 2 jx(f )j2 df

Z 1 21 1 Z 1 12
! $t $f = ;1 jt x(t)j dt
2 ;1 jx_ (t)j dt
kxk;2
2 2

Using Schwartz' inequality, and integrating by parts yield:

$t $f  41 Gabor's relation


If the signal x(t) is centered around to , $t is dened about to , and the same if X (f ) is
centered around f = fo . The Gabor's relation remains the same.

Consequences of the uncertainty principle


 $t $f   ! $t and $f cannot both approach zero.
4
1

 In comunications the bandwidth required to transmit a message increases when the speed
of information transmission is increased.
B. Boashash, Signals in Linear Systems
50 Chapter 3. Spectral Representation of Signals
 There is an analogy with quantum mechanics when the position and velocity of a particle
obeys:

$p $x  h2 Heisenberg's uncertain principle

 If we use another denition of bandwidth and duration, we will still obtain an inequality
but with a dierent limit.
 Given a linear lter h(t) $ H (f ), $t represents the memory of the lter, $f is its
bandwidth ! $t $f  41 .
 Case of random process: Rx ( ) $ Sx (f ). $ t represents the statistical memory on
\correlation time" and f is the spectral bandwidth.

$ $f  41

white process ! $f ! 1 and $ ! 0


 This is a consequence of the property (t) $ 1.
 The equality $t $f  41 is obtained for the Gaussian signal:

t2
x(t) = k1 e;  2 $ k e;r2 2f 2
2

This signal is its own Fourier Transform.


Important property: The Gaussian signal has the minimum required bandwidth for a
given duration of message.
3.4.4 Time{Frequency Distributions
 the spectrogram
 the WVD
 the B distribution
 instantaneous frequency

B. Boashash, Signals in Linear Systems


3.4. Time{Varying Spectra 51

Table 3.2: Some selected unilateral Laplace transform pairs.

Signal Transform ROC


u(t) 1 <fsg > 0
s
u(t) ; u(t ; a) 1 ; e;as <fsg > 0
s
(t) 1 for all s
(t ; a) e;as for all s
tnu(t) n! <fsg > 0
sn+1 , n = 1
2
: : :
e;at u(t) 1 <fsg > ;a
s+a
tn e;at u(t) n! <fsg > ;a
(s + a)n+1
cos(2ft)u(t) s <fsg > 0
s + 42 f 2
2

sin(2ft)u(t) 2f <fsg > 0


s + 42 f 2
2

cos(2ft)]2 u(t) s2 + 82 f 2 <fsg > 0


s(s2 + 162 f 2 )
sin(2ft)]2 u(t) 8 2 f 2 <fsg > 0
s(s2 + 162 f 2 )
e;at cos(2ft)u(t) s+a <fsg > 0
(s + a)2 + 42 f 2
e;at sin(2ft)u(t) 2f <fsg > 0
(s + a)2 + 42 f 2
t cos(2ft)u(t) s2 ; 42 f 2 <fsg > 0
(s2 + 42 f 2)2
t sin(2ft)u(t) 4fs <fsg > 0
(s + 42 f 2)2
2

B. Boashash, Signals in Linear Systems


52 Chapter 3. Spectral Representation of Signals

Table 3.3: Some selected properties of the Laplace transform.

Signal Transform
x(t) X (s)
X
N X
N
n xn(t) n Xn(s)
n=1 n=1
x(t ; t0 ) X (s)e;t0 s
es0 tx(t) X (s ; s0 )
x(t) 1X s
 
dx(t) sX (s) ; x(0; )
Z t dt 1 X (s)
;
x( )d s
0

tx(t) ; dXds(s)
x(t) cos(2f0 ) 1 X (s ; j 2f ) + X (s + j 2f )]
2 0 0

x(t) sin(2f0 ) 1 X (s ; j 2f ) ; X (s + j 2f )]


2j 0 0

x(t)  h(t) X (s) H (s)


x(0+ ) lim sX (s)
s!1
lim x(t)
t!1
lim sX (s)
s!0

B. Boashash, Signals in Linear Systems


Chapter 4

System Analysis in the Frequency


Domain
4.1 Introduction
The Fourier transform is a very important tool that nds extensive application in communication
systems, signal processing, control systems, and many other areas of physical and engineering
disciplines. For example, amplitude modulation and frequency multiplexing use the Fourier
transform in the analysis and design of communication systems. The design of lters that are
employed in communication systems cannot be done without using the Fourier transform.

4.2 An Example of Frequency Domain System Analysis


The reception of radio signals requires separating the signal of interest from other signals in
adjacent bands. A band{pass lter is used to perform this task.
X(f)
B(f)

fc f

Figure 4.1: Band-pass ltering of desired signal.


This system may be described as

X (f ) - B (f ) - Y (f )
(selected radio station)

Y (f ) = X (f )
B (f )
53
54 Chapter 4. System Analysis in the Frequency Domain
Practical systems may be described and realised in the frequency domain. Since the Fourier
transform allows convolution of signals and systems to be represented as multiplications in
frequency, it is eective to analyse signals and implement systems in the frequency domain.

4.3 Transfer Function


x(t) h(t) Z1
- LSI system -y (t) = x( )h(t ; )d
X (f ) H (f ) ;1

The properties of the Fourier Transform led to:


y(t) = x(t)  h(t) ! Y (f ) = X (f )
H (f )
H (f ) is called the transfer function of the system and is identical to the frequency response
function that we have seen before. As H (f ) is, in general, complex, we can write:
H (f ) = jH (f )j ej
(f )
jH (f )j is the amplitude response function of the network, and "(f ) is its phase response function.
If h(t) is a real time function, it follows that H (;f ) = H (f ) , which results in
jH (f )j = jH (;f )j
"(f ) = ;"(;f ):
Example 4.3.1 Obtain the transfer function of the low-pass RC lter.
R
d
i(t)
- d
6 6

x(t) C y(t)

d d

Consider this circuit as a system relating input and output


x(t) = R i(t) + y(t)
i(t) = C dydt(t)
RC dydt(t) + y(t) = x(t)
;1 < t < 1 (4.1)
This expression is the signal-system (input-output) mathematical relationship in dierential
equation form. If the Fourier transform of each term exists, it follows
(1 + j 2f RC ) Y (f ) = X (f )
B. Boashash, Signals in Linear Systems
4.3. Transfer Function 55
or
Y (f ) =
H (f ) = X 1 1
(f ) 1 + j 2fRC = 1 + j ff0
The frequency f0 = 2RC
1
is called the 3 dB frequency.
The amplitude response of this system is
" f 2 #; 12
jH (f )j = 1 + f0
and is shown in Figure 4.2.
|H(f)|
1
0.707

-f 0 0 f0 f

Figure 4.2: Amplitude response of a low-pass lter.


The phase response is "(f ) = ; tan;1 ff0 and is shown in Figure 4.3. Let us now nd the

/2

/2

Figure 4.3: Phase response of a low-pass lter.


response of the system to the input. If this was done in the time domain, it would be necessary
to solve the dierential equation (4.1) for x(t) = (t), thus determine h(t), since y(t) = h(t) for
x(t) = (t) and then convolve x(t) and h(t).
For an input signal
x(t) = A e;t u(t)
>0
the Fourier transform is given by
X (f ) =  +Aj 2f

B. Boashash, Signals in Linear Systems


56 Chapter 4. System Analysis in the Frequency Domain
The Fourier transform of the output due to this input is
Y (f ) = ( + j 2f A=RC
)(1=RC + j 2f )
which can written as
A
1 1

Y (f ) = RC ; 1 1=RC + j 2f ;  + j 2f
The inverse Fourier transform is, if  RC 6= 1
h i
y(t) = RCA ; 1 e; RCt ; e;t u(t)
and if RC = 1, then by taking the limit when RC ! 1, we get
 t  t 1
y(t) = A RC e; RC u(t)
 = RC
Example 4.3.2 Low-pass RC lter with pulse input. Let
 t ; T=2 
x(t) = Arect T = A u(t) ; u(t ; T )]
The step response of this system is
 
as(t) = 1 ; e; RCt u(t)
The output y(t) consists of the dierence of two steps.
8
>
< 0
  t<0
y(t) = > ;
A 1 e; RCt

  0 t T
: A exp t;T

; RC ; exp ; RCt 
t>T
The result is plotted in gure 4.4 for several values of RC T together with jX (f )j and jH (f )j.
T
The parameter RC is proportional to the ratio of the 3 dB frequency of the lter to the spectral
width (T ;1 ) of the pulse. When the lter bandwidth is large compared with the spectral width of
the input pulse, the input is essentially undistorted by the system. For 2f0 =T  1, the system
distorts the input signal spectrum and the output does not resemble the input.

4.4 Steady{State System Response to Sinusoidal Inputs by Means


of the Fourier Transform
Let us study the relation between the Fourier transforms of a xed, linear system in response
to a periodic waveform with period T = 1=f0 , where f0 is the fundamental frequency.
1
X
x(t) = Xn ej 2nf0t (4.2)
n=;1
Because
ej 2nf0t ! (f ; nf0)
B. Boashash, Signals in Linear Systems
4.4. Steady{State System Response 57

abs[X(f)] abs[X(f)]
abs[H(f)] abs[H(f)]

Input pulse

T/RC=10

T/RC=0.5

T 2T t

Figure 4.4: Time and frequency domain plots.

we can nd the Fourier transform of x(t) in terms of the Fourier coecients Xn , namely
1
X
X (f ) = Xn (f ; nf0):
n=;1
Then
1
X
Y (f ) = H (f )
X (f ) = Xn H (f ) (f ; nf0):
n=;1
Using the relationship H (f ) (f ; nf0) = H (nf0) (f ; nf0), we have
1
X
Y (f ) = Xn H (nf0) (f ; nf0 )
n=;1
Therefore
1
X
y(t) = Xn H (nf0 )e+j2f0 nt:
n=;1
Writing Xn and H (nf0 ) in terms of magnitude and phase,
1
X
y(t) = jXnjjH (nf )j ej nf0t
0
2 +
n +
H (nf0 )

n=;1
We note that the nth spectral component of the input, Xn , appears at the output with amplitude
attenuated, or amplied by the amplitude response function jH (nf0 )j.
Example 4.4.1
jH (f )j
6

-f
;B2
B
2

B. Boashash, Signals in Linear Systems


58 Chapter 4. System Analysis in the Frequency Domain
Consider a system with amplitude and phase{response functions given by
f 
jH (f )j = krect B
"H (f ) = ;2fto
A lter with this transfer function is called an ideal low-pass lter. Its output in response to
x(t) = A cos(2f0 t + "0 ) is obtained from the expression of y(t) (see above) by noting that
X1 = 21 A ej
0 = X; 1 and Xn = 0
jnj 6= 1
Thus the output is

y(t) = 0
f0 > B=2
kA cos 2f0 (t ; t0 ) + "0 ]
f0 B=2
An ideal low{pass lter completely rejects all spectral components with f > B=2 (cut-o fre-
quency), and passes all input f below this cuto B=2, except that their amplitudes are multiplied
by a constant k and delayed in time by t0 (or phase shifted by ;2t0 f0 ).

4.5 Ideal Filters


Any LSI system performs some kind of ltering. It is often convenient to work with idealised
lters having amplitude-response functions which are constant within the passband and zero
elsewhere. The passband pof a lter is dened as the frequency range where its amplitude
response is greater than 1= 2 (3 dB bandwidth).
There are three types of ideal lters, low{pass, bandpass, and high{pass lter. In each case,
we assume a phase{shift function which is a linear function of frequency in the passband. The
impulse response of any lter can be found by inverse Fourier transform of its transfer function.
For example, for the ideal low{pass lter, hLP (t) is given by
Z1
hLP (t) = H (f )ej 2ftdf
Z;1
1 f 
= krect B e;j2ft0 ej 2ft df
;1
= 2Bk sinc2B (t ; t0 )]
We see that hLP (t) is nonzero for t < 0. Therefore an ideal low{pass lter is non-causal.
An ideal bandpass lter is causal. Its impulse response is given by
hBP (t) = 2kB sincB (t ; t0 )] cos 2f0 (t ; t0 )
The amplitude response is depicted in Figure 4.5.
An ideal high{pass lter is also causal, whose impulse response is given by
hHP (t) = k (t ; t0 ) ; 2Bksinc2B (t ; t0)]
The amplitude response of the lter is depicted in Figure 4.6.
B. Boashash, Signals in Linear Systems
4.6. Fourier Transforms of Periodic Signals and Sampling 59

jH (f )j
6

k
- f
;f 0 f0
Figure 4.5: Transfer function of an ideal bandpass lter.

jH (f )j
6

- f
; B
2
B
2

Figure 4.6: Transfer function of an ideal high{pass lter.

4.6 Fourier Transforms of Periodic Signals and Sampling


1
X
For a periodic signal x(t) with Fourier series Xn ej 2f0nt , where T0 = f10 is the period, we
n=;1
have
1
X 1
X
x(t) = Xn ej 2f0nt ! X (f ) = Xn (f ; nf0 )
n=;1 n=;1

Either representation (Fourier series or Fourier transform) contains the same information about
x(t), and either result can be used to plot the spectra of a signal.
To obtain x(t), we can consider it as the result of convolving the ideal sampling waveform
with a pulse{type signal p(t). The signal p(t) is an energy signal of limited duration such that

p(t) = 0
jtj  T2 with T Ts :

Then x(t) is a periodic power signal


" 1 # 1
X X
x(t) = (t ; mTs )  p(t) = p(t ; mTs)
m=;1 m=;1
B. Boashash, Signals in Linear Systems
60 Chapter 4. System Analysis in the Frequency Domain
Thus
( 1 )
X
X (f ) = F (t ; mTs ) P (f )
m=;1
1 X1  n

= T P (f ) f;T
s n=;1 s
1
X 1  n
  n

= P T f;T :
n=;1 Ts s s
Example 4.6.1 Consider the periodic signal
1
X
x(t) = (t ; nT )
n=;1
which has period T . To nd Fourier transform, we rst have to compute the Fourier-series
coe cients
1 Z
x(t)e; T dt = 1
j 2nt
c = n T T
<T>
since x(t) = (t) in any given interval of length T . Thus, the impulse train has the Fourier
series representation
1 1 j2nt
X
x(t) = e T :
n=;1 T
Then, the Fourier transform of the impulse train can be given by
X1  
X (f ) = 1 f;n
T n=;1 T
That is, the Fourier transformation of a sequence of impulses in the time-domain yield a sequence
of impulses in the frequency domain.
Example 4.6.2 Given the transform pair, $(t=T ) $ T sinc2 (fT ), obtain the Fourier Trans-
form of the periodic triangular waveform, given in Figure 4.7.
6x(t)
1
;@ ;@ ;@
; @ ; @ ; @
@ @ ; ; ;
@;
;
@;
@
@;
@
@; -
;9 ;6 ;3 3 6 9 t

Figure 4.7: Periodic triangular signal.

1
X 1 
X 
xp(t) = $(t=3)  (t ; 6n) ) Xp (f ) = F f$(t=3)g  16 f ; n6
n=;1 n=;1
B. Boashash, Signals in Linear Systems
4.6. Fourier Transforms of Periodic Signals and Sampling 61
X1   X1 n  n
Xp (f ) = 3sinc2 (3f ) 61 f ; n6 = 12 sinc2
2 f;6
n=;1 n=;1

B. Boashash, Signals in Linear Systems


62 Chapter 4. System Analysis in the Frequency Domain

B. Boashash, Signals in Linear Systems


Chapter 5

Introduction to Discrete{time Signal


Processing
5.1 Analog to Digital Conversion
A digital signal can be formed from an analog signal using an Analog{Digital Converter.

;! Sampler ;! Quantiser ;! Encoder ;!


continuous{time discrete{time discrete{time digital output
continuous{amplitude continuous{amplitude discrete{amplitude signal
analog/input signal signal signal

Figure 5.1: ANALOG{DIGITAL CONVERTER

5.1.1 Sampling
To sample a continuous{time signal x(t) at a discrete number of points, t = nT , where T is the
sampling period (time between samples), and n is an integer that establishes the time position
of each sample.
To extract samples of x(t), the sampling switch closes every T seconds. This yields:
 a value of x(t) when the switch is closed
 0 when the switch is open.
A useful sampling process requires that x(t) can be reconstructed from the samples.
Reconstruction of x(t) from its samples
The sample signal xs(t) is written:

xs(t) = x(t)p(t)
where p(t), called the sampling function, models the action of the sampling switch (see
gure 5.2). p(t) is periodic ! it can be represented by its Fourier series:
1
X
p(t) = Cn ejn2fst (5.1)
n=;1

63
64 Chapter 5. Introduction to Discrete{time Signal Processing
Samples of x(t)

t
0 T 2T 3T 4T 5T 6T 7T

Samples of a waveform

Switch closes at t =nT


x (t) xs (t)

Sampling device

Figure 5.2: Sampling of a continuous signal

where Cn is the nth Fourier coecient of p(t) and is given by

Z
Cn = T1
T
2
p(t)ejn2fst dt (5.2)
;2
T

In equations 5.1 and 5.2, fs is the fundamental frequency of p(t), which is the sampling
frequency, and is given by

fs = T1 hertz

x (t) xs (t)

p(t)

Figure 5.3: A typical sampling process

B. Boashash, Signals in Linear Systems


5.1. Analog to Digital Conversion 65
p(t)

t
0 T 2T 3T 4T 5T 6T 7T

Figure 5.4: The waveform p(t). Ideally, p(t) should be a periodic set of delta functions

Since xs(t) is the product of x(t) and p(t), we have (using eq.5.1)

1
X
xs(t) = Cn x(t)ejn2fs t
n=;1
The Fourier transform of xs (t) is dened by
Z1 X
1
Xs (f ) = Cnx(t)ejn2fs te;j 2ft
;1 n=;1
Interchanging the order of integration and summation yields (as per Fubini's theorem)

1
X Z1
Xs(f ) = Cn x(t)e;jn2(f ;nfs)t dt
n=;1 ;1
From the denition of the Fourier transform,
Z1
X (f ; nfs) = x(t)e;jn2(f ;nfs)t dt
;1
Thus the Fourier transform of the sampled signal can be written
1
X
Xs(f ) = Cn X (f ; nfs)
n=;1
The spectrum of the sampled continuous{time signal, x(t), is composed of the spectrum of
x(t) plus the spectrum of x(t) translated to each harmonic of the sampling frequency. Each of the
translated spectra is multiplied by a constant, given by the corresponding term in the Fourier
series expansion of p(t).
If the sampled signal is ltered by the reconstruction lter, the output of the lter is, in the
frequency domain C0 X (f ) and the time domain signal is C0 x(t).
Important: Sampling in the time domain , periodising in the frequency domain.
B. Boashash, Signals in Linear Systems
66 Chapter 5. Introduction to Discrete{time Signal Processing

X(f)

- fn 0 fn f
Xs(f)
fs =1/T, fs > 2 fn

0f f
- fs - fn n fs

H(f)

Low-pass filter
fn < fc < fs - fn

- fc 0 fc f

Y(f)

- fn 0 fn f
Figure 5.5: Exact recovery of a continuous{time signal from its samples using an ideal low{pass
lter

B. Boashash, Signals in Linear Systems


5.1. Analog to Digital Conversion 67
In gure 5.5, x(t) is assumed band limited. In other words, X (f ) is assumed zero for jf j  fh.
The condition to recover x(t) from xs (t) is then:

fs ; fh  fh
or
fs  2fhHz
Sampling Theorem:
A band limited signal x(t), having no frequency components above fh hertz, is completely
specied by samples that are taken at a uniform rate greater than 2fh hertz. Then, the time
between samples is no greater than 1/2fh seconds.
The frequency 2f is called the Nyquist rate.
In a practical situation, a signal cannot be strictly band limited, because it is time limited.
However, in all practical signals, there is some frequency beyond which the energy is negligible.
! denition of bandwidth.
Another source of error is the nonexistence of ideal reconstruction lters. Practically, we
need to sample at a higher frequency than the Nyquist rate.

5.1.2 Periodic Sampling: a tutorial


Given the importance of the sampling process in modern engineering, this material is presented
a second time as revision and example suited for a class tutorial.
Let xc (t) be a continuous-time signal and its Fourier transform be denoted by Xc (f )
Z +1
Xc(f ) = xc(t)e;j 2ft dt :
;1
The signal xc (t) is obtained from Xc (f ) by
Z +1
xc(t) = Xc(f )e+j2ft df :
;1
The typical method of obtaining a discrete-time representation of a continuous-time signal is
through periodic sampling, wherein a sequence of samples x(n) is obtained from the continuous-
time signal xc (t) according to
x(n) = xc(t)jt=nT = xc(nT )
n = 0
1
2
: : :
T > 0:
Herein, T is the sampling period, and its reciprocal, fs = 1=T , is the sampling frequency in
samples per second. This equation is the input-output relationship of an ideal A/D converter.
It is generally not invertible since many continuous-time signals can produce the same output
sequence of samples. Fortunately, this ambiguity can be removed by restricting the class of
input signals to the sampler.
It is convenient to represent the sampling process in two stages, as shown in Figure 5.6. This
consists of an impulse train modulator which is followed by conversion of the impulse train into
a discrete sequence.
B. Boashash, Signals in Linear Systems
68 Chapter 5. Introduction to Discrete{time Signal Processing

xc (t) -  - G - x(n) = xc(nT )


6x t s( )

p(t)

Figure 5.6: Continuous-Time to digital converter.

The system G in Figure 5.6 represents a system that converts an impulse train into a discrete-
time sequence. This is often referred to as a zero{order hold (for more deatils, see 5] p. 520).
The modulation signal p(t) is a periodic impulse train,
1
X
p(t) = (t ; nT )

n=;1
where (t) is called the Dirac delta function. Consequently, we have
1
X 1
X 1
X
xs (t) = xc(t)
p(t) = xc(t)
(t ; nT ) = xc(nT ) (t ; nT ) = x(n) (t ; nT ):
n=;1 n=;1 n=;1
The Fourier transform of xs (t)
Xs (f ), is obtained by convolving Xc (f ) and P (f ). The Fourier
transform of a periodic impulse train is a periodic impulse train, i.e.,
1 1
X k 1 1
X
P (f ) = T k=;1 (f ; ) = T (f ; kf ):
T k=;1 s

Since
Xs(f ) = Xc(f )  P (f )

it follows that
1
X    1
X
Xs(f ) = T1 Xc j f ; Tk = T1 Xc (jf ; kjfs ) :
k=;1 k=;1
Figure 5.7 depicts the frequency domain representation of impulse train sampling. Figure
5.7(a) represents a band-limited Fourier transform where the highest non-zero frequency com-
ponent in Xc (f ) is at fB . Figure 5.7(b) shows the periodic impulse train P (f ), Figure 5.7(c)
shows Xs (f ), which is the result of convolving Xc (f ) with P (f ). From Figure 5.7(c) it is evident
that when
fs ; fB > fB
or fs > 2fB
the replicas of Xc (f ) do not overlap, and therefore xc (t) can be recovered from xs(t) with an
ideal low pass lter. However, in Figure 5.7(d) the replicas of Xc (f ) overlap, and as a result
the original signal cannot be recovered by ideal low pass ltering. The reconstructed signal is
related to the original continuous-time signal through a distortion referred to as aliasing.
If xc (t) is a band-limited signal with no components above fB Hz, then the sampling fre-
quency has to be equal to or greater than 2fs Hz. This is known as the Nyquist criterion and
2fs is known as the Nyquist frequency.

B. Boashash, Signals in Linear Systems


5.1. Analog to Digital Conversion 69

(a) X c (f a )
1

f f fa
B B

(b)
P(fa )
1

T

2 f s fs f s 2 f s fa

(c)
1 X s (f a )
a
T

fB fs fa

(d)
X s (f a )
1

T

fs fa

Figure 5.7: Eect in the frequency domain of sampling in the time domain: (a) Spectrum of
the continuous-time signal, (b) Spectrum of the sampling function, (c) Spectrum of the sampled
signal with fs > 2fB , (d) Spectrum of the sampled signal with fc < 2fB .

B. Boashash, Signals in Linear Systems


70 Chapter 5. Introduction to Discrete{time Signal Processing
5.1.3 Reconstruction of Band{Limited Signals
A bandlimited signal x(t) can be reconstructed from xs (t) through an ideal low{pass lter with

bandwidth

B  fh
B fs ; fh
Reconstruction ltering:
If x(n) is the input to an ideal low pass lter with frequency response Hr (f ) and impulse
response hr (t), then the output of the lter will be
1
X
xr (t) = x(n)h(t ; nT ) :
n=;1
where T is the sampling interval. The reconstruction lter commonly has a gain of T and a
cuto frequency of fc = fs =2 = 1=2T . This choice is appropriate for any relationship between
fs and fB that avoids aliasing, i.e., so long as fs > 2fB .
1
X Ideal Reconstruction
 Filter
x(nT ) (t ; nT ) ! T
j f j
H (f ) = 0
s elsewhere 0 :5fs ! x(t)
n=;1

Reconstruction ltering
The impulse response of such a lter is given by:
h (t) = sin (t=T ) :
r t=T

hr (t) = Tfs sinfftst = sincfs t


s
The response to x(0) (t) is x(0)h(t) = x(0)sincfst.
The response to the nth sample x(nT ) (t ; nT ) is:

x(nT )h(t ; nT ) = x(nT )sincfs(t ; nT )


Using the linearity property, we have
1
X
x(t) = x(nT )sincfs(t ; nT )
n=;1
Consequently, the relation between xr (t) and x(n) is given by

xr (t) =
1
X t ; nT )=T ]

x(n) sin(t(;
n=;1 nT )=T
with xc (nT ) = x(n).
B. Boashash, Signals in Linear Systems
5.1. Analog to Digital Conversion 71

sample of x (t)
x (t)

(n-2)T (n-1)T nT (n+1)T (n+2)T

Figure 5.8: Ideal reconstruction

Example 5.1.1
x(t) = 6 cos 2(50)t
We consider this signal samples at 70 and 140Hz. The Nyquist Frequency is 100Hz.

X (f ) = 3 (f ; 50) + 3 (f + 50)
Spectrum of the sampled signal:

1
X
X (f ) = 3fs  (f ; 50 ; nfs) + (f + 50 + nfs)]
n=;1
The assumed reconstruction lter is an ideal low{pass lter with a bandwidth of 0.5fs and
an amplitude response of T .

Example 5.1.2 We consider a nonperiodic signal for x(t) so that X (f ) has a real, continuous
spectrum.
The highest frequency is 50Hz ! the minimum acceptable sampling frequency is 100Hz.
Again, we assume sampling frequencies of 70 and 140Hz.
For a sampling frequency of 70Hz aliasing occurs. For fs = 140 Hz, there is no spectral
overlap since fs > 100Hz.
The reconstruction lter is an ideal low{pass lter with an amplitude response of T and a
bandwidth of 0.5fs .
The output spectrum of the reconstruction lter is shown for f = 70Hz and fs = 140Hz. The
impact of aliasing is clear.

B. Boashash, Signals in Linear Systems


72 Chapter 5. Introduction to Discrete{time Signal Processing

X (f)

Spectrum of x (t)

-50 0 50 f

Reconstruction filter
Xs(f)
transfer function
fs = 70 Hz

0 20 50 90 120 160 190 f


Xs(f)
Reconstruction filter
transfer function fs = 140 Hz

0 50 90 190 f
Xs(f)H (f) Xs(f)H (f)

-20 0 20 f -50 0 50 f
Output of reconstruction Output of reconstruction
filter with sampling frequency filter with sampling frequency
equal to 70 Hz. equal to 140 Hz.

Figure 5.9: Sampling a single sinusoid at 70 and 140 Hz

B. Boashash, Signals in Linear Systems


5.1. Analog to Digital Conversion 73

X (f)

Spectrum of x (t)

-50 0 50 f

Reconstruction filter Xs(f)


transfer function Spectrum of sampled signal
with sampling frequency equal to 70 Hz.

-50 0 50 70 140 210 f

Reconstruction filter Xs(f)


transfer function Spectrum of sampled signal
with sampling frequency equal to 140 Hz.

-190 -140 -90 -50 0 50 90 140 190 f

Xs(f)H (f) Xs(f)H (f)

-35 0 35 f -50 0 50 f

Output of reconstruction filter with Output of reconstruction filter with


sampling frequency 70 Hz. sampling frequency 140 Hz.

Figure 5.10: Sampling of a non{periodic signal at 70 and 140 Hz

B. Boashash, Signals in Linear Systems


74 Chapter 5. Introduction to Discrete{time Signal Processing
5.1.4 Quantising and Encoding
Quantization Encoded
level number output

7 111 S

6 110

5 101

4 100

3 011

2 010

1 001

0 000
t
0 T 2T 3T 4T

Figure 5.11: Quantising and encoding


The process of quantising and encoding is illustrated above. To quantise a sample value is
to round o the sample value to the nearest of a nite set of permissible values. To encode is to
represent each of the permissible values by a digital word.
If q is the number of quantising levels and if n is the digital word length, we have:

q = 2n
In the above example, we have 8 levels, and each level is uniquely specied by a three{bit
word.
Transfer t characteristic:
X
xQ(t) = ms(s (x ; ms)
m

xQ

s
with nullarea

-3s/2 -s/2

s/2 3s/2 x (t)

-s

Figure 5.12: Transfer characteristic (with null error)

B. Boashash, Signals in Linear Systems


5.1. Analog to Digital Conversion 75

X
xQ(t) = (ms + 2s )(s (x ; ms ; 2s )
m

xQ

without nullarea

s/2
x (t)
-s s

-s/2

Figure 5.13: Transfer characteristic (without null error)


A quantication without error is such that xQ = x.
Quantising process errors: (see gure 5.14)

nx(x) = x ; xQ

 The maximum error induced by quantising a sample is  1


2 S , where S is the width of a
quantising level.
 The minimum error is zero.
 If we assume that the number of quantising levels q is large, then S results in a small
value.
! Then, for most quantising levels, the signal x(t) will be nearly linear within the quantising
level. For the system of interest, it is the samples that are quantised and not the continuous{time
signal x(t).
For band limited signals, sampling and reconstruction can be accomplished without error for
band limited signals.
! Then the quantiser is the only error source in our A=BD converter model. This error is
shown in the above gure.
The spectral density of this noise is in general constant in the bandwidth ; f2s
f2s ], fs =
1
T = sampling ratio.
Ts s
The mean{square error is given by:
Z t1 Z t1
E = 21t 2()d = t1 2 ()d
1 ;t1 1 0

B. Boashash, Signals in Linear Systems


76 Chapter 5. Introduction to Discrete{time Signal Processing

Center of
quantization
S
level

t
-t 1 0 t1

a) Quantizing

(t)

S/2

-t 1
t
0 t1

-S/2

b) Quantizing error

2
(t)

2
S /4

t
-t 1 0 t1

c) Square of quantizing error

Figure 5.14: Calculation of quantising error

The error is given by: (t) = 2St t


Z t1 S 1
1 ( 2t )2 2 d = S12 independent of time.
2
Then, E = t
1 0 1
E is interpreted as a noise power,
The quality criterion of an A=D converter is in the SNR (Signal to Noise Ratio).
Signal to Noise Ratio: SNR = SIGNAL POWER
NOISE POWER
The eect of quantising can then be interpreted as the sum of the original signal with a uniform
random noise with mean{square value E fn2 g = S12 , and mean value E fng = 0. The spectral
2

F F
density of the noise n(t) is in general constant in the bandwidth ; 2s
2s , where Fs = T1 .
s
We dene D = A=D converter dynamic range

D = max x(t)] ; min x(t)]


B. Boashash, Signals in Linear Systems
5.2. Digital Processing of Continuous-Time Signals Using the DTFT 77
as there are q = 2n quantising levels, the width of quantising level is:

S = 2Dn = D
2;n
which yields

E = D12
2;2n
2

In calculating the signal power, the assumption is usually made that the signal power at the
output of the A=D converter is equal to the signal power at the input of the A=D converter.
Example 5.1.3 We want to sample and quantise the signal

x(t) = A cos 2f0 t


The signal power is
Z Tp
1 (a
cos 2f0 t)2 dt = A2
2
x =T
2
p 0

where Tp is the period.


The dynamic range is 2A.
Thus, the noise power is: E = 412A2
2;2n = A32
2;2n

SNR = AE=2 = 32
22n
2

SNR(dB) = 10
log10 SNR = 10
log10 23 + 20n
log10 2
SNR = 1:176 + 6:02n
Thus, the signal{to{noise ratio at the output of the A=D converter increases by approximately
6 dB for each added bit of word length.
! importance of using a large word length.
We obtain a similar result with any test signal: the bias, here 1:76, will change with the
shape of the test signal, but in general, the SNR increases 6:02dB for each increment in the
word length.

5.2 Digital Processing of Continuous-Time Signals Using the


DTFT
Processing of continuous-time signals is often carried out by discrete-time processing of sequences
obtained by sampling. It is remarkable that under reasonable constrains a continuous-time signal
can be adequately represented by samples.
B. Boashash, Signals in Linear Systems
78 Chapter 5. Introduction to Discrete{time Signal Processing
SNR

50

25

n
4 8 12

Figure 5.15: SNR versus word length


5.2.1 Discrete{time Processing of Continuous{time Signals.
A continuous-time signal is processed by digital signal processing techniques using analog-to-
digital (A/D) and digital-to-analog (D/A) converters before and after processing. This concept
is illustrated in Figure 5.16. The low pass lter limits the bandwidth of the continuous-time
signal to reduce the eect of aliasing. The spectral characteristics (magnitude) of the low pass
lter is shown in Figure 5.17.
Analog Analog Digital Processed
(or Continuous - Low pass - A/D - Signal - D/A - Contontinuous-Time
-Time) Signal Filter Processing Signal

Figure 5.16: Discrete-time processing of continuous-time signals.

6H (f ) Processed
ya (t) Signal
- -
-
-fB 0 fB f

Figure 5.17: Continuous-Time Filter.


The advantages of the digital approach compared with the continuous-time approach are:
1. Flexibility, if we want to change the continuous-time lter because of change in signal and
noise characteristics, we would have to change the hardware components. Using the digital
approach, we only need to modify the software.
B. Boashash, Signals in Linear Systems
5.2. Digital Processing of Continuous-Time Signals Using the DTFT 79
2. Better control of accuracy requirements. Tolerances in continuous-time circuit components
make it extremely dicult for the system designer to control the accuracy of the system.
3. The signals can be stored without deterioration or loss of signal quality.
4. Lower cost of the digital implementation.

5.2.2 The Discrete{Time Fourier Transform


Denition of the DTFT
Many discrete time sequences x(n) can be obtained by appropriately combining complex expo-
nentials of the form X (f )ej 2fn . The relationships between x(n) and X (f ) are given by
1
X
X (f ) = x(n)e;j 2fn=fs
n=;1
and Z fs=2
x(n) = f1 X (f )ej2fn=fs df

s ;fs=2
where fs is the sampling frequency. For fs = 1 Hz which will be assumed in the sequel unless
other specied, we have Z 1=2
x(n) = X (f )ej2fndf :
;1=2
X (f ) is called the discrete time Fourier transform of x(n). Conversely, x(n) is the inverse Fourier
transform of X (f ). Sometimes it is useful to consider the equation for X (f ) as an operator that
transforms the sequence into a function, and we will refer to the Fourier transform operator as
1
X
Ffx(n)g = x(n)e;j 2fn = X (f ):
n=;1
Consequently x(n) will be obtained as
x(n) = F ;1 fX (f )g:
X (f ) is in general complex and is a function of a continuous variable f . X (f ) is always periodic
with period of 1, i.e., X (f ) = X (f + r), for all r 2 Z because it is the FT of a sampled signal.
This may be veried as follows:
1
X
X (f + r) = x(k)e;j 2fk e;j 2r = X (f )

k=;1
using the fact that ej 2r = 1 for all integer r.
We dene a sequence x(n) to be an eigenfunction of a system T if Tx(n)] = kx(n), for
any scalar k. Clearly, ej 2fn is the eigenfunction of any LSI system, and the scaling factor k
is H (f ), the Fourier transform of the system impulse response h(n). This concept is shown in
Figure 5.18, where the scaling factor H (ej! ) is called the transfer function of the system or the
frequency response function.
B. Boashash, Signals in Linear Systems
80 Chapter 5. Introduction to Discrete{time Signal Processing
x(n) = ej 2fn - LSI - H (f )
ej2fn
h(n)

Figure 5.18: LSI system excited by x(n) = ej 2fn .

5.2.3 Convergence of the DTFT


The function X (f ) is said to converge uniformly when X (f ) is nite and
X
N
lim
N !1
x(n)e;j2fn = X (f ) for all f :
n=;N
X (f ) uniformly converges only for absolutely summable sequences (sucient condition). The
sequence n u(n) is not an absolutely summable sequence for jj > 1, and therefore its Fourier
transform does not exist. Some sequences are not absolutely summable but are square summable,
i.e., 1
X
jx(n)j2 < 1 :
n=;1
Such sequences can be represented by a Fourier transform if we are willing to relax the condition
of uniform convergence of the innite sum dening X (f ). Specically, in this case we have mean-
square convergence with 1
X
X (f ) = x(n)e;j2fn
n=;1
and
X
M
XM (f ) = x(n)e;j2fn
n=;M
then Z 1=2
lim
M !1 ;1=2
jX (f ) ; XM (f )j df = 0
2

In other words the error jX (f ) ; XM (f )j may not approach zero at each f as M ! 1, but the
total energy in the error does.
Example 5.2.1 Consider an ideal low pass lter with cut o frequency fc. Its transfer function
is given by 
H (f ) = 1
jf j < fc

0
fc < jf j 1=2 :
The impulse response of the lter is given by
Z fc
h(n) = ej 2fndf
;fc
= j 21n ej 2fc n ; e;j 2fc n ]
h(n) = sin(2nfc n) ;1<n<1
B. Boashash, Signals in Linear Systems
5.2. Digital Processing of Continuous-Time Signals Using the DTFT 81
This is the impulse response of a non casual lter because h(n) 6= 0 for n < 0. Also, h(n) is not
absolutely summable since
1 sin(2f n)
X c e;j 2fn
n=;1 n
does not converge uniformly for all values of f . However, because these functions have such an
important role in digital ltering, it is accepted as a valid Fourier transform pair.

Exercise: Plot the following transfer function


XM
sin(2fc n) e;j 2fn
HM (f ) = n
n=;M

and let M increase. Does HM (f ) approach H (f )?

Example 5.2.2 Consider the sequence x(n) = 1 for all n. This sequence is neither absolutely
nor square summable. However it is useful to assign the periodic impulse train
1
X
X (f ) = (f + k)

k=;1

where (f ) is Dirac's impulse function to be its Fourier transform. One can check that indeed
this result is valid if X (f ) is substituted in
Z 1=2
x(n) = X (f )ej 2fn df:
;1=2

5.2.4 Properties of the DTFT.


 Linearity: Due to the linear property of the Fourier transform, one can easily show that
ax (n) + bx (n) ! aX (f ) + bX (f )
1 2 1 2

 Time and frequency shift:


x(n ; n0) ! X (f )e;j2fn0
holds because
X X
x(n ; n0 )e;j 2fn = x(k)e;j2f (k+n0 ) = e;j2fn0
X (f )
n k

 Convolution: 1
X
x(n) ? h(n) = x(k)h(n ; k) ! X (f )
H (f )
k=;1
B. Boashash, Signals in Linear Systems
82 Chapter 5. Introduction to Discrete{time Signal Processing
which results from the fact that
1 X
X
Ffx(n) ? h(n)g = x(k)h(n ; k)e;j2fn
n=;1 k
X X
= x(k) h(n ; k)e;j2fn
k n
X X
= x(k) h(u)e;j2f (u+k)
u
X X
= x(k)e j2fk
h(u)e;j2fu
;
k u
= X (f )H (f )

 Values at the origin: Z


x(0) = 1=2X (f )df
;1=2
1
X
X (0) = x(n)
n=;1

 Parseval's theorem :
1
X Z
x(n)
y(n) = 1=2X (f )
Y (f ) df
n=;1 ;1=2
which simplies to
1
X Z
jx(n)j 2
=
;1=2
1=2jX (f )j2 df
n=;1
This result holds because
1
X 1 Z
X Z
x(n)y(n) = 1=2X (f )ej 2fn df 1=2Y (ej 2 ) e;j 2 nd
n=;1 n=;1 ;1=2 ;1=2
Z 1=2 Z 1=2 " 1 #
X
= X (f )Y ( ) ( ; f + k) dfd
;1=2 ;1=2 k=;1
Z
= 1=2X (f )Y (f ) df
;1 =2

 If x(n) is absolutely summable, then X (f ) uniformly converges.


The results of this section and a number of Fourier transform pairs are summarised in Table 5.1
and Table 5.2. Many results obtained in Table 5.2 are based on the following formulae
1
X
an = 1 ;1 a if jaj < 1
n=0
B. Boashash, Signals in Linear Systems
5.2. Digital Processing of Continuous-Time Signals Using the DTFT 83
For example, if we were to compute the Fourier transform of n e;j 2fn , then
1
X 1
X
ne;j2fn = e;j 2f ]n
n=;1 n=0
= 1 ; e1;j 2f
jj < 1:
Note that the Fourier transform of n u(n) does not exist when jj  1, since the summation
does not converge.

B. Boashash, Signals in Linear Systems


84 Chapter 5. Introduction to Discrete{time Signal Processing

Table 5.1: Discrete{Time Fourier transform theorems.

Sequence Fourier Transform


ax(n) + by(n) aX (f ) + bY (f )
x(n ; nd ) (nd integer) e;j 2fnd X (f )
ej 2f0 nx(n) X (ej 2(f ;f0 ) )
x(;n) X (;f )
X (f ) if x(n) real
nx(n) j dX (f )
2df
x(n) ? y(n) X (f )Y (f )
Z 1=2
x(n)y(n) X ( )Y (ej2(f ; )d
;1=2
;jnx(n) dX (f )
2df
Symmetry properties:
x(n) X (;f )
x(n) real X (f ) = X (;f )
x(n) real XR (f ) andjX (f )j are even
XI (f ) and argX (f ) are odd

x(;n) X (;f )
x(n) real and even X (f ) real and even
x(n) real and odd X (f ) pure imaginary and odd
Parseval's Theorem
1
X Z 1=2
jx(n)j = ; = jX (f )j d!
2 2

n ;1
= 1 2
Z =
X1
x(n)
y(n) = X (f )
Y (f )df
1 2

n=;1 ;= 1 2

B. Boashash, Signals in Linear Systems


5.2. Digital Processing of Continuous-Time Signals Using the DTFT 85

Table 5.2: Discrete{Time Fourier transform pairs.

Sequence Fourier Transform


(n) 1
(n ; n0 ) e;j2fn0
X1
1, (;1 < n < 1) (f + k)
k=;1
an u(n), (jaj < 1) 1
1 ; ae;j 2f 1
1 1 X
u(n) 1 ; ae;j 2f + 2 k=;1 (f + k)
(n + 1)an u(n) (jaj < 1) 1
n
(1 ; ae;j 2f )2
r sin(2fp (n + 1)) u(n) (jrj < 1) 1
sin(2fp ) 1 ; 2r cos(2 f
8 p e ; j 2f ) + r2 e;j 22f
sin(2fc n) < 1
jf j < fc
n X ( f ) = : 0
fc < jf j 1=2
8
< 1
0 n M f (M + 1)=2] e;j2fM=2
x(n) = : x sin2sin(2 f=2)
0
otherwise
1
X
ej 2f0m (f ; f0 + k)
k=;1
1 X1
cos(2(fn + )) 2 ej 2 (f ; f0 + k) + e;j 2 (f + f0 + k)
k=;1

B. Boashash, Signals in Linear Systems


86 Chapter 5. Introduction to Discrete{time Signal Processing
5.3 The z{Transform
The z{transform is the discrete{time counterpart of the Laplace transform and the generalization
of the discrete{time Fourier transform.
5.3.1 De
nition of the z{Transform
A suitable model of a sampled signal is
1
X
xs(t) = x(t) (t ; nT )
n=;1
1. As (t ; nT ) equals zero except at the sampling instants, t = nT ,
x(t) can be replaced by x(nT ).
2. We assume x(t)  0, t < 0, to establish a time reference.

1
X
xs(t) = x(t) (t ; nT )
n=;1
Taking the Laplace Transform
Z 1X
1
Xs(s) = x(nT ) (t ; nT )e;st dt
0 n=0
Interchanging integration and summation
1
X Z1
Xs(s) = x(nT ) (t ; nT )e;st dt
n=0 0

Using the property of (t)


1
X
Xs(s) = x(nT )e;snT
n=0
Let us dene z = esT . Then
1
X
X (z ) = x(nT )z ;n
n=0
X (z ) is called the z {transform of the sequence of samples, x(nT ). Given a sequence of
samples, a given sample can be represented by an ordered pair of numbers.
 one number in the pair represents the value of the sample.
B. Boashash, Signals in Linear Systems
5.3. The z{Transform 87
 the other number is used to specify the occurence time of the sample
In the summation of X (z ), the coecient, x(nT ) represents the sample value.
z;n denotes that the sample occurs in sample periods after t = 0.
z = esT is only a short notation for the T time shift operator.
Example 5.3.1
12:8z ;16 represents a sample, having value 12:8, which occurs 16 sample periods after t = 0. We
have:

s =  + j! ! z = e T ej!T
Stable functions of the Laplace variable s must not have any poles in the right{half s{plane
or on the jw{axis. We can then derive a condition for the z {transform.
The magnitude of z is given by: jz j = e T
 The right{half s{plane,  > 0, corresponds to jzj > 1.
 The left{half s{plane,  < 0, corresponds to jzj < 1.
Im

z=e j

Unit circle z-plane

Re

Figure 5.19: Unit circle in the complex z -plane.

Example 5.3.2
The unit pulse sequence is dened by the sample values
 
x(nT ) = 1
n = 0 4
= (n)
0
n 6= 0
Its z {Transform is then:

X (z ) = 1 + 0
z;1 + 0
z;2 +

Thus

X (z ) = 1
B. Boashash, Signals in Linear Systems
88 Chapter 5. Introduction to Discrete{time Signal Processing
1

-3 -2 -1 0 1 2 3 n

Figure 5.20: The unit pulse.

Important Result: The unit pulse sequence plays the same role in discrete{time systems that
the unit impulse function plays in analog systems. For example, the sequence p(n) in Figure
5.21 can be expressed as
p(n) = a;1 (n + 1) + a1 (n ; 1) + a3 (n ; 3) :
Any sequence x(n) can be represented as a linear combination of delayed unit sample sequences,

6p(n)
s
a;1 a1 s a3 s
-n
;1 1 3
Figure 5.21: The sequence p(n).

1
X
x(n) = x(k) (n ; k)

k=;1
as shown in Figure 5.22.
Example 5.3.3
The unit step sample sequence is dened by the sample values
X (nT ) = 1
n0
= 0
otherwise
Its z {Transform is then:
1
X
X (z ) = z;n
n=0
We know that for jxj < 1
1
X
xn = 1 ;1 x
n=0
B. Boashash, Signals in Linear Systems
5.3. The z{Transform 89
6x(n)
t
t t
t t t t t t t t t
-
1 n

t t t t
6 x(k)
(n ; k )

t t t t t t t t t
-
n k

Figure 5.22: Discrete Time Signal.


1

-3 -2 -1 0 1 2 3 4 5 6 7 n

Figure 5.23: The unit step sequence.

Thus,
X1
X (z ) = z ;n = 1 ;1z ;1
jzj > 1
n=0
The unit step sample sequence is often denoted u(n). The sequence u(n) is related to (n)
by
X
n
u(n) = (k )

k=;1
or alternatively
1
X
u(n) = (n ; k ) :
k=0
Conversely, the impulse sequence can be expressed as the rst backward dierence of the unit
step sequence, i.e.,
X
n ;1
nX
(n) = u(n) ; u(n ; 1) = (k) ; (k )
k=;1 k=;1

B. Boashash, Signals in Linear Systems


90 Chapter 5. Introduction to Discrete{time Signal Processing
Example 5.3.4
The unit exponential sequence is dened by the sample values

x(nT ) = e;nT
 > 0
n  0

x(n)

Figure 5.24: Exponential Sequence with 0 <  < 1.


Its z {transform is:
1
X 1 ;
X 
X (z ) = e;nT z;n = e;T z;1 n
n=0 n=0
P xn = 1
using 1 1;x , for jxj < 1 yields, for x = e
;T z ;1
n=0

X (z) = 1 ; e;1T z ;1
jz j > e;T
jzj > e;T denes;the
T
ROC: region of convergence.
Given  and T , e is a constant k.
Then

X (z) = 1 ; k1
z;1
jzj > k
Example 5.3.5 x(n) = an u(n).
The z -transform of x(n) is given by
1
X 1
X 1 ;
X 
X (z) = an u(n)z ;n = anz ;n = az;1 n :
n=;1 n=0 n=0
The series converges if jz j > jaj. Then,
X (z ) = 1 ; 1az ;1 = z ;z a
jzj > jaj :
B. Boashash, Signals in Linear Systems
5.3. The z{Transform 91
The Fourier transform of x(n) converges only if
1
X
janu(n)j < 1

n=;1
which requires that
1
X
jajn < 1

n=0
which is the case for jaj < 1. For jaj = 1, we have x(n) = u(n) and X (z ) = z;z 1 , jz j > 1, as
discussed in the previous example. A typical ROC is shown in Figure 5.25. Outside the ROC,
i.e., Rx+ < jz j < Rx; , X (z ) does not converge and does not exist.

Im

R x+

Re

R x-

Figure 5.25: Region of Convergence ROC: Rx; < jz j < Rx+.

5.3.2 Properties of the z{Transform


1. Linearity:
1
X
Ax1 (nT ) + Bx2 (nT )] z ;n = AX1 (z ) + BX2 (z )
n=0
2. Initial Value Theorem:
x(0) = zlim
!1 X (z )
Proof 5.3.1
X (z ) =
P1 x(nT )z;n = x(0) + P1 x(nT )z;n
n=0 n=0
limz!1 X (z ) = x(0) + 0
Final Value Theorem:
x(1) = zlim
!1
(1 ; z ;1 )X (z )

B. Boashash, Signals in Linear Systems


92 Chapter 5. Introduction to Discrete{time Signal Processing
Proof 5.3.2 It is assumed that X (z ) decomposes as

X (z ) = 1 ;kz;1 + G(z)

where all the poles of G(z ) lie inside the unit circle ! x(1) = k

5.3.3 Inverse z{Transform


The z {Transform of a sample sequence is, by denition:

X (z ) = x(0) + x(T )z;1 + x(2T )z ;2 +

If X (z ) is in this form, the sample values, x(nT ), can be determined by inspection. This
form is easily obtained by long division when X (z ) is expressed as a ratio of polynomials in z .
Example 5.3.6

X (z ) = (z ; 1)(zz ; 0:2)
2

Let us determine its inverse.


First, we develop X (z ) as follows:

X (z) = z2 ; 1:z2z + 0:2 = 1 ; 1:2z;11 + 0:2z;2


2

= 1 + 1:21z ;;11+:2z1:24z+;02:+
;1 2z ;2
1:248z ;3
! X (z) = 1 + 1:2z;1 + 1:24z;2 + 1:248z;3 +

and we have: x(0) = 1


x(T ) = 1:2
X (2T ) = 1:24
x(3T ) = 1:248

From initial value theorem:


x(0) = zlim
!1 X (z ) = 1
From nal value theorem:
x(1) = zlim z ; 1 X (z) = lim z = 1:25
!1 z z!1 z ; 0:2
Although this method doesn't yield the general term, it is sometimes sucient, especially
when you have an idea of the behaviour of the signal.
Another method is the partial{fraction{expansion method. The idea is to write X (z ) into a
form that can be inverse z {transformed by using a table of transforms.
B. Boashash, Signals in Linear Systems
5.3. The z{Transform 93
Example 5.3.7

X (z ) = z 2 ; 1:z2z + 0:2 = (z ; 1)(zz ; 0:2)


2 2

we write:

X (z ) = z = k1 + k2
z (z ; 1)(z ; 0:2) z ; 1 z ; 0:2
where:

k1 = limz!1 (z ; 1) Xz(z) = 1:25


k2 = limz!0:2(z ; 0:2) z = ;0:25
X (z )

Thus,
25z ; 0:25z = 1:25 ; 0:25
X (z ) = 1z:; 1 z ; 0:2 1 ; z ;1 1 ; 0:2z ;1
We nd the inverse transform of X (z ) by using the list of transforms1 . The rst term has
k = 1 ! it's a unit step sample sequence. The second term has k = 0:2 which is a sampled
exponential.
Thus x(nT ) = 1:25 ; 0:25(0:2)n
n  0
The values of x(nT ) are: x(0) = 1:25 ; 0:25(0:2)0 = 1
x(T ) = 1:25 ; 0:25(0:2)1 = 1:2
x(2T ) = 1:25 ; 0:25(0:2)2 = 1:24
x(3T ) = 1:25 ; 0:25(0:2)3 = 1:248
Delay Operator Delay Operation is very important in digital systems. If the time sequence
fx(nT )g is delayed by k sample periods, the eect in the z{domain is to multiply X (z) by z;k .
Proof 5.3.3 X (z) = P1n=0 x(nT )z;n
The z {Transform of the sequence fx(nT ; kT )g is:

1
X
x(nT ; kT )z;n
n=0
P1
Let m = n ; k ! m=;k x(mT )z ;m;k
As x(mT ) is assumed zero for m < 0, we have:

1
X
z ;k x(mT )z;m = z ;k X (z)
m=;k
1 Gabel and Roberts, Signals in linear systems, 2nd edition, p212 and 213

B. Boashash, Signals in Linear Systems


94 Chapter 5. Introduction to Discrete{time Signal Processing
5.4 Digital Systems
A dierential equation can be used to model an analog system ! a dierence equation is used
to model a discrete{time system.
Example 5.4.1
Consider an analog system described by the dierential equation:

dy + a y(t) = b x(t)
dt
x(t) ! System ! y(t)
The output y(t) can be expressed as:
Zt Zt
y(t) = b x()d ; a y()d
;1 ;1
This equation tells us thatRthe present value of the system output, y(t), is a function of all
t
R t of the input, b ;1 x()d, and is also a function of all past values of the system
previous values
output, b ;1 y()d.
5.4.1 Dierence Equations
A linear digital system operates in the same way: the present output, y(nT ), is computed using
the present input x(nT ), past inputs x(nT ; kT ), and past system outputs y(nT ; kT ).
The general dierence equation for this processor is:
y(nT ) = Lox(nT ) + L1 x(nT ; T ) + L2x(nT ; 2T ) +

+Lr x(nT ; rT ) ; k1 y(nT ; T ) ; k2 y(nT ; 2T ) ;

;kmy(nT ; mT )
The processor generates an output by weighting the present input, the past n inputs, and
the past m outputs. The processor is illustrated in the Fig. 5.26.

Figure 5.26: Linear digital signal processor

The analysis problem is usually a problem of determining the system output given the system
input and a specication of the system. The most convenient method is to specify the coecients
of the dierence equation.
Another problem is the synthesis of digital lters in which the problem is to determine the
coecients of the dierence equation in order to perform some specied task.
The rst step is to solve the dierence equation. We can write:
y(nT ) + k1 y(nT ; T ) + k2 y(nT ; 2T ) +

+ km y(nT ; mT )
= Lo x(nT ) + L1 x(nT ; T ) + L2 x(nT ; 2T ) +

+ Lr x(nT ; rT )
B. Boashash, Signals in Linear Systems
5.4. Digital Systems 95
Now, we z {transform both sides using the time delay operator.
Z x(nT ; kT )] = z;k X (z)
Y (z ) + k z ; Y (z) + k z; Y (z ) +

+ km z;mY (z)
1
1
2
2

= Lo X (z ) + L z ; X (z ) + L z ; X (z ) +

+ Lr z ;r X (z )
1
1
2
2

That is:
Y (z )(1 + k1 z ;1 + k2 z;2 +

+ km z;m )
= X (z )(Lo + L1 z ;1 + L2 z ;2 +

+ Lr z ;r )
It follows:
H (z) = L1o++kLz1;z 1 ++kLz2;z2 ++

++k Lzr;z m
;1 ;2 ;r
1 2 m
H (z) = XY ((zz)) is called the pulse transfer function.
Relation input{output in the z {domain: Y (z ) = H (z )X (z ).
We have seen that the z {Transform of a unit pulse is unity. Thus, if a digital system has
a unit pulse input, the system output is: Y (z ) = H (z ). The inverse z {Transform yields:
y(nT ) = h(nT ).
f h(nT ) g is called the unit pulse response. In digital systems, it is the equivalent of the
impulse response in analog systems.
These systems are shift invariant since a time shift of the input time shifts the output without
changing the shape of the output.
If h(nT ) is zero for n < 0, the system is called a causal system.
5.4.2 Discrete Convolution
Formulation
Let us determine the time{domain equivalent of

Y (z) = H (z )X (z)
By denition, we have:

X (z ) = x(0) + x(T )z;1 + x(2T )z;2 +

and

H (z ) = h(0) + h(T )z;1 + h(2T )z;2 +

where x(iT ) = ith value of the system input and h(iT ) = ith value of the unit pulse response.
The three preceding equations yield:
 
Y (z ) = x (0) + x(T )z;1 + x(2T )z ;2 +


h(0) + h(T )z;1 + h(2T )z;2 +

B. Boashash, Signals in Linear Systems


96 Chapter 5. Introduction to Discrete{time Signal Processing
That can be written in groupings terms with the same z ;n :
Y (z) = x(0)h(0) + x(0)h(T ) + x(T )h(0) +

] z;1 +
+ x(0)h(2T ) + x(T )h(T ) + x(2T )h(0) +

] z ;2
We note that: if a term multiplies z ;n , the sum of the arguments of x and h is nT .
Thus, the general term is:

x(0)h(nT ) + x(T )h(nT ; T ) +

+ x(nT ; T )h(T ) + x(nT )h(0)] z ;n


Since

Y (z ) = y(0) + y(T )z;1 + y(2T )z ;2 +

+ y(nT )z;n +

we have: y(0) = x(0) h(0)


y(T ) = x(0) h(T ) + x(T ) h(0)
y(2T ) = x(0) h(2T ) + x(T ) h(T ) + x(2T ) h(0)
..
.
y(nT ) = x(0) h(nT ) + x(T ) h(nT ; T ) +

+x(nT ; T ) h(T ) + x(nT ) h(0)


X
n X
n
! y(nT ) = x(mT )h(nT ; mT ) = h(mT )x(nT ; mT )
m=0 m=0
This denes the discrete convolution.

Denition
Example 5.4.2
See Fig. 5.27.
We shall use discrete convolution to compute the output of a discrete processor whose input
x(nT ) and unit pulse response h(nT ) are shown in Fig. 5.27.

X
n
y(nT ) = x(mT )h(nT ; mT )
m=0

Example 5.4.3 To compute the relationship between the input and the output with a transversal
lter, h(;k) is rst plotted against k h(;k) is simply h(k) reected or \ipped" around k = 0,
as shown in Figure 5.27. Replacing k by k ; n, where n is a xed integer,
P leads to a shift in the
origin of the sequence h(;k) to k = n. y(n) is then obtained by the x(k)h(n ; k). See Figure
5.27.

B. Boashash, Signals in Linear Systems


5.4. Digital Systems 97
6x(k) 6h(k)

t t t
t
3
1
2
- 1 t 2
-
k k
6y(n) 7 t t
6h(;k) 6
4 t
2 t t
1 1 t
- -
k n

Figure 5.27: Convolution operation.

Steady{State Frequency Response of a Linear Digital System The frequency steady{


state response of a xed, linear, discrete{time system is determined by placing the complex
sampled sinusoid

x(nT ) = Aej(!nT +)


on the input of a discrete{time system.

! Linear, xed discrete{time system !


x(nT ) = Aej (!nT +) y(nT ) = Aej (!nT +)
h(n)
H (z )
G(F ) = frequency response = H (ej!T ) we replace z by ej!T  B = amplitude
scaling = phase shift.
Using the discrete convolution theorem, we can write:
X
n
y(nT ) = A h(mT )ej !(n;m)T +]
m=0 " n #
j (!nT +)
X ; j!mT
y(nT ) = Ae h(mT )e
m=0

For n suciently large, the terms h(nT ) of the unit pulse response are negligible ! the
steady{state response has been reached. The term in brackets is then a complex function
depending only on the input frequency !.
B. Boashash, Signals in Linear Systems
98 Chapter 5. Introduction to Discrete{time Signal Processing
We denote it
X
n
G(!) = h(mT )e;j!mT
m=0
! y(nT ) = G(!)Aej (!nT +)
We know that:
1
X
H (z) = h(mT )z;m
m=0
Comparing the two equations G(!) and H (z ), it follows:
; 
G(!) = H ej!t
The steady{state response is easily obtained from the pulse transfer function by replacing z
in the pulse transfer function with ej!t .
Summary:
 previous result
 importance of the ROC.

B. Boashash, Signals in Linear Systems


Bibliography
1] Boashash, B., ed. Time-Frequency Signal Analysis. Methods and Applications, Longman
Cheshire, 1992
2] Boashash, B., Powers, E.J. and Zoubir, A.M., eds. Higher-Order Statistical Signal Processing,
Longman, 1995
3] Kamen, E.W. Introduction to Signals and Systems, 2nd Edition, MacMillan, 1990.
4] Oppenheim, A.V. and Wilsky, A.K. Signals and Systems, Prentice-Hall, 1983.
5] Oppenheim, A.V., Wilsky, A.K., and Nawab, H. Signals and Systems, Prentice-Hall, 1996.
6] Soliman, S.S. and Srinath, M.D. Continuous and Discrete Signals and Systems, Prentice-
Hall, 1990.
7] Ziemer, R.E., Tranter, W.H. and Fannin, D.R. Signals and Systems, 3rd Edition, MacMillan,
1993.

99

Das könnte Ihnen auch gefallen