Lecture 2 and a part of Lecture 3: Introduction

to signals

TUT/Pretoria
February 9th , 2015

Contents

1 Foreword 1

2 classification of signals 2

3 Piecewise-defined signals 5

4 Size (strength) of a signal 6

4.1 Signal energy (E f ) . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
4.2 Signal power (Pf ) . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
4.3 Use of the measures of signal size . . . . . . . . . . . . . . . . . 7
4.4 Energy signal and Power signal . . . . . . . . . . . . . . . . . . . 8
4.5 Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

5 Signal operations 13
5.1 Time shifting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
5.2 Time inversion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
5.3 Time scaling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
5.4 combined operations . . . . . . . . . . . . . . . . . . . . . . . . . 16
5.5 Amplitude transformations . . . . . . . . . . . . . . . . . . . . . 17

6 Signal models 19
6.1 Unit step function u(t) . . . . . . . . . . . . . . . . . . . . . . . . 19
6.2 Unit impulse function δ(t) . . . . . . . . . . . . . . . . . . . . . . 23
6.3 Complex exponential function est . . . . . . . . . . . . . . . . . . 26

7 Even and odd functions 27

7.1 Some properties of even/odd functions . . . . . . . . . . . . . . 27
7.2 Finding even and odd components of a signal . . . . . . . . . . 28

1 Foreword
Definition 1.1. A signal is a set of information or data. A Signal is mathemat-
ically defined as a function of one or more than one independent variables:

1
 2. classification of signals

f (t), f ( x, y), etc. Independent variables can be time, distance, spatial coordi-
nates, etc.

• Examples:
20

15

10

−5

−10

−15

−20
0 20 40 60 80 100 120 140 160 180 200
n(sample number)

Example of an Electroencephalogram (EEG) signal. Example of Magnetic Resonance

Imaging (MRI) image of a part of a
brain.

• Other signals: speech signal, television signal, daily closing prices of a

stock market, monthly sales of a corporation, temperature, number of
schools per provinces, etc.

• Signal processing IV deals with signals that are functions of time: Signals
that are function of one independent variable, time. Such signals are
sometimes referred to as time series.

2 classification of signals

Signals can be distinguished in different ways. Here we consider only some

distinctions suitable to the scope of the study.

• Continuous-time signal: A signal that is specified for every value of time t

(See Figure 1.5 a).

• discrete-time signal: A signal that is specified only at discrete values of t

(See Figure 1.5 b).
B[discrete-time and continuous-time referring to the nature of a signal
along the time (horizontal) axis].

• Analogue signal: A signal whose amplitude can take on any value in a

continuous range [i.e., an analogue signal can have an infinite number
of values in a given continuous range].

2
 • Digital signal: A signal whose amplitude can take on only a finite number
of values. [A digital signal whose amplitude can take on M values is an
M-ary signal of which binary (M = 2) is a special case. For example, the
transmission of 3 bits at a time (simultaneously) over a communication
channel is possible with an 8-ary signal (i.e., 8 different signal levels)].
Analogue signal can be converted into a digital signal (Analog-to-digital
(A/D) conversion) through quantization.
BAnalogue signal and digital signal referring to the nature of the signal
amplitude (nature of the signal along the vertical axis).
8 8

7 7

6 6

5 5
f(t)

f(t)

4 4

3 3

2 2

1 1

0 0
−1 −0.8 −0.6 −0.4 −0.2 0 0.2 0.4 0.6 0.8 1 −1 −0.8 −0.6 −0.4 −0.2 0 0.2 0.4 0.6 0.8 1
t (sec) t (sec)

Analogue, continuous-time signal. Digital, continuous-time signal.

8 8

7 7

6 6

5 5
f(t)

f(t)

4 4

3 3

2 2

1 1

0 0
−1 −0.8 −0.6 −0.4 −0.2 0 0.2 0.4 0.6 0.8 1 −1 −0.8 −0.6 −0.4 −0.2 0 0.2 0.4 0.6 0.8 1
t (sec) t (sec)

Analogue, discrete-time signal (plot using stem). Digital, discrete-time signal (plot us-
ing stem).
The upper left figure being the first figure, all the remaining three figures
are obtained from this first figure. See also Figure 1.5 in the textbook for
another example.

3
 2. classification of signals

• Periodic signal A signal f (t) is periodic if for some positive constant T0 :

f (t) = f (t + T0 ) for all t (periodicity condition).
Examples of periodic signals:
2π
1. f (t) = sin(ω0 t), T0 =
ω0
2. Figure 1.6

The period of f (t) is the smallest value of T0 that satisfies the periodicity
condition.
Properties:

1. A periodic signal f (t) remains unchanged when time-shifted by

one period (i.e., a periodic signal must start at t = −∞ and continue
forever).
2. A periodic signal f (t) can be generated by periodic extension of
any segment of f (t) of duration T0 (the period) [see Figure 1.7]

• Aperiodic signal: A signal that is not periodic

• Causal signal: A signal is causal if f (t) = 0 for t < 0

• Noncausal signal: A signal is noncausal if f (t) ̸= 0 for t < 0.

Periodic signals are always noncausal but a noncausal signal is not nec-
essarily periodic.
Example of causal and noncausal signals:
1 1

0.8 0.8

0.6 0.6
f(t)

f(t)

0.4 0.4

0.2 0.2

0 0

−0.2 −0.2
−1 −0.8 −0.6 −0.4 −0.2 0 0.2 0.4 0.6 0.8 1 −1 −0.8 −0.6 −0.4 −0.2 0 0.2 0.4 0.6 0.8 1
t (sec) t (sec)

Causal signal. Noncausal signal.

• Energy signal and Power signal: The difference between the two signals is
explained in the following section (section on Size of signal).

• A deterministic signal: A signal whose physical description is known com-

pletely either in a mathematical form or graphical form. Such signals can
be approximated by mathematical models (e.g., sin(t), et ) and the values
of these signals can be predicted precisely.

4
• A random signal: A signal whose values cannot be predicted precisely but
are known only in terms of probability description, such as mean value,
mean squared value (or variance) and so on. Such signal can in general
be a stochastic signal (or stochastic process) with a probability distribu-
tion defined at each time instant and a correlation defined across differ-
ent time instants (e.g., speech, transient signals); however, it can also be
a signal (random in time) whose values over a given time interval have
a specific probabilistic value or frequency spectrum (e.g., white noise
signal).
[Signal processing IV deals only with deterministic signals.]

3 Piecewise-defined signals

• Some signals, referred to as piecewise-defined signals are most conveniently

modelled by a set of equations with each equation valid over a differ-
ent time interval (or in general over a subdomain of the signal domain).
Piecewise is actually a way of expressing the signal, rather than a char-
acteristic of the signal itself.

• An example:



 t2 + 4t + 4 −2 ≤ t < −1

0.16t2 − 0.48t + 0.36 −1 ≤ t < 1.5
f (t) =

 e−0.5t 1.5 ≤ t < 3

0 otherwise(or t ≥ 3 and t < −2)

The graph of f (t) is the following:

1.2

0.8

0.6
f(t)

0.4

0.2

−0.2
−5 −4 −3 −2 −1 0 1 2 3 4 5
t (sec)

An example of MATLAB function to plot the above graph:

5
4. Size (strength) of a signal

f = (t.ˆ2 + 4 ∗ t + 4). ∗ ((t >= −2)&(t < −1))

+(0.16 ∗ t.ˆ2 − 0.48 ∗ t + 0.36). ∗ ((t >= −1)&(t < 1.5))
+ exp(−0.5 ∗ t). ∗ ((t >= 1.5)&(t < 3));

• It is commonly to assume right continuous (use of t ≥ a and t < b where

a and b are the left and right limits of an interval, respectively)

• A simply-defined signals is modelled by a single equation for all time.

• Some signals are compactly defined by a single equation but are more
conveniently used in analysis if a piecewise definition is employed. For
example when f (t) = e−|t| is used in integral, it can be broken into two
parts:

{
et t<0
f (t) =
e−t t≥0

and then add integral from −∞ to 0 and the integral from 0 to ∞.

4 Size (strength) of a signal

The signal size is a measure on a signal f (t) stretching over a certain time
interval and such measure is embodied in one number. Noting that the signal
values can vary within the time interval and the signal size considers then
both the signal values and the length of the interval. One possible measure of
∫ can be the area under f (t) [distinguish geometric
the signal size ∫ area, which
is given by | f (t)|dt and algebraic area, wich is given by f (t)dt].
Two measures of signal size are considered in this study, namely, signal
energy and signal power.

4.1 Signal energy (E f )

∫∞
Ef = | f (t)|2 dt for a general complex-valued signal f (t). For a real signal,
−∞
∫∞
Ef = f 2 (t)dt. Energy is then the area under f 2 (t) [ f 2 (t) is always positive
−∞
for a real signal].
E f is not the actual energy of the signal and it is useful for characterizing
and comparing signals. For example in an electric circuit, if the signal f (t)
represents the voltage across an electrical resistor or the current through the

6
 4.2. Signal power (Pf )

resistor, E f will have to be divided by the resistance value of the resistor if

f (t) is the voltage or multiplied by the resistance value of the resistor if f (t)
is the electrical current; i.e., the actual signal energy, contrary to E f , depends
on the load [Energy dissipated in a resistor of resistance R is equals to V It =
V2 I2
R = R ].
Signal energy must be finite (E f < ∞) for it to be a meaningful measure
of the signal size. A necessary condition for E f to be finite is that the signal
value f (t) tends to 0 as |t| tends towards ∞ (see Figure 1.1 (a)).

4.2 Signal power (Pf )

Signal power is the ratio of the energy in a given time interval to the length of
the interval (i.e., time average of the energy in the interval) and is given by:
T
∫2
1
Pf = lim | f (t)|2 dt for a general complex-valued signal f (t). In case
T →∞ T
− T2
T
∫2
1
f (t) is a real signal, Pf = lim f 2 (t)dt. Pf is the mean squared value of
T →∞ T
− T2
f (t) and the square root of Pf gives the root mean square (rms) value of f (t)
(i.e., signal power of f (t) is the square of the rms value of f (t)).
Similar to signal energy, Pf size is also not the actual power of the signal
and its use has the same purpose as E f .
The mean(average) of | f (t)|2 averaged over a large time interval approach-
ing infinity exists if f (t) is either periodic, or though with an infinity energy,
has a finite variance.

4.3 Use of the measures of signal size

Both E f and Pf are not correct dimensionally (i.e., not expressed in Joules or
Watt, respectively): The units of E f and Pf depend on the nature of the signal
f (t). For example: If f (t) is the voltage signal, E f has units of Volt2 .second
(volts squared-seconds) and Pf has units of Volts2 (volts squared), while if
f (t) is a electrical current signal, unit for E f is A2 .second (amperes squared-
seconds) and unit for Pf is A2 (amperes squared).
The two signal sizes can be used in different task, such as:

• Comparing signals:

1. Approximate a signal f (t) by another signal g(t). The error in the

approximation is e(t) = f (t) − g(t). The energy or power of e(t)
is an indicator of the goodness of the approximation (i.e., the two
quantities provide a quantitative measure of determining the close-
ness of the approximation).
2. During transmission of message signals over a channel in commu-
nication systems, message signals are corrupted by unwanted sig-

7
4. Size (strength) of a signal

nals (noise). The quality of the received signal is judged by the rela-
tive sizes of the desired signal and the unwanted signal (noise): The
ratio of the message signal and noise signal powers (Signal to noise
power ratio) is a good indication of the received signal quality.

• Signal energy or signal power in different components of a single signal

indicates the relative significance of the components.

4.4 Energy signal and Power signal

• Energy signal: A signal whose energy is finite (i.e., E f < ∞)

• Power signal: A signal whose power is finite and nonzero (i.e., Pf < ∞
and Pf ̸= 0, otherwise 0 < Pf < ∞).
A power signal must necessarily have infinite duration; otherwise its
power which is the energy averaged over an infinitely large interval will
not approach a (nonzero) limit.
Periodic signals for which the area under | f (t)|2 (or under f 2 (t) for real-
valued signals) over one period is finite are power signals; however not
T
∫2
1
all power signals are periodic. For periodic signals Pf = | f (t)|2 dt
T
− T2
for a general complex-valued signal f (t) with T being the period of the
T
∫2
1
periodic signal. In case f (t) is a real signal, Pf = f 2 (t)dt with T
T
− T2
being the period of the periodic signal.

• Since the averaging is over an infinitely large interval, a signal with finite
energy (E f < ∞) has zero power (Pf = 0) and a signal with finite power
(Pf < ∞) has infinite energy (E f = ∞). Therefore, a signal cannot be both
an energy signal and a power signal [if it is one, it cannot be the other];
however there are signals that are neither energy nor power signals. For
example:

1. A signal that contains infinite energy (E f = ∞) and infinite power

(Pf ) [in other words, its energy and power do not exist]: Unit ramp
signal, i.e.,

{
t t≥0
f (t) =
0 t<0

A ramp signal increases indefinitely as t → ∞: Its energy increases

indefinitely as t → ∞ and its power increases for each increase in
the length of a time interval.
A ramp signal has a slope and a y-intercept.

8
 4.5. Examples

2. A signal that contains infinite energy (E f = ∞) and zero power

(Pf = 0):

{
t−0.1 t≥1
f (t) =
0 t<0

4.5 Examples

Which one of the following signal is power signal or energy signal? Motivate.

(a) f 1 (t) = e−2|t|

π
(b) f 2 (t) = sin(2t + )
3

• Solution

0.8

0.6
f(t)

e−2|t|

0.4

0.2

−1 −0.8 −0.6 −0.4 −0.2 0 0.2 0.4 0.6 0.8 1

t (sec)
(a)

9
4. Size (strength) of a signal

∫∞
E f1 = | f 1 (t)|2 dt
−∞
∫∞
E f1 = |e−2|t| |2 dt
−∞
∫∞
= e−4|t| dt
−∞
∫0 ∫∞
= e4t dt + e−4t dt
−∞ 0
1 1 −4t ∞
= [e4t ]0−∞ + [ e ]0
4 −4
1 1 1 1
= (1 − 0) + (0 − 1) = +
4 −4 4 4
1
=
2
Noting that from the second line, we can also solve the problem knowing
∫∞ ∫∞
that E f1 = |e−4t |dt = 2 e−4t dt
−∞ 0

=⇒ Finite signal energy

∫
T/2
1
Pf1 = lim f 12 (t)dt = 0
T →∞ T
− T/2

=⇒ Zero signal power

f 1 (t) is an energy signal.

10
 4.5. Examples

0.8

0.6
sin(2t+ π/3)
0.4

0.2
f(t)

−0.2

−0.4

−0.6

−0.8

−1

−π −2π/3 −π/2 −π/6 0 π/3 π/2 5 π/6 π

t (sec)
(b)

∫∞
E f2 = f 22 (t)dt = ∞
−∞

π
The graph of sin2 (2t + ) is shown below:
3

1 sin2(2t+ π/3)

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

−π −2π/3 −π/2 −π/6 0 π/3 π/2 5 π/6 π

t (sec)

π
The area between sin2 (2t + ) and the t axis for −∞ < t < ∞ is infinity
3
=⇒ Infinite signal energy

11
4. Size (strength) of a signal

∫
T/2
1
Pf2 = lim | f 2 (t)|2 dt
T →∞ T
− T/2
∫
T/2
1 π 2
= lim | sin(2t + )| dt
T →∞ T 3
− T/2
∫
T/2
1 π
= lim sin2 (2t + )dt
T →∞ T 3
− T/2

The signal is periodic with a period equals to 2π, then taking T = 2π

we have:
∫π
1 π
Pf 2 = sin2 (2t + )dt
2π 3
−π

1
Given that sin2 θ = (1 − cos(2θ )):
2
∫π
1 1 2π
Pf 2 = (1 − cos(4t + ))dt
2π 2 3
−π

Given that cos(θ1 + θ2 ) = cos θ1 cos θ2 − sin θ1 sin θ2

∫π ( )
1 2π 2π
Pf 2 = 1 − cos(4t) cos( ) − sin(4t) sin( ) dt
4π 3 3
−π

12
 √
2π 1 2π 3
Given that cos( ) = − and sin( ) =
3 2 3 2
∫π
( √ )
1 1 3
Pf 2 = 1 + cos(4t) − sin(4t) dt
4π 2 2
−π
[ √ ]π
1 1 3
= t + sin(4t) + cos(4t)
4π 8 8
−π
√ √
1 3 3
= [(π + 0 + ) − (−π − 0 + )]
4π 8 8
1 1
= (2π ) =
4π 2

=⇒ Finite signal power and nonzero

f 2 (t) is a power signal.

• Do example 1.1

• Do example 1.2

• Do Exercise E1.1

• Do Exercise E1.2

• Do Exercise E1.3

• Do Exercise E1.4: Show that e at is neither an energy nor a power signal

for any real value of a. However if a is imaginary, it is a power signal
with power Pf = 1 regardless of the value of a.

• A complex-valued signal such as f (t) = 2e3jt can also be assessed for

power and energy.

5 Signal operations
In this section, three time transformations (namely, Time shifting, Time in-
version(folding: Time reversal) and Time scaling) and three amplitude trans-
formations (namely, amplitude shifting, amplitude inversion and amplitude
scaling) are described.

5.1 Time shifting

Given a signal f (t), its time-shifted g(t) is given by g(t) = f (t − T ) where T
is a constant.

(1) T > 0, the shift is to the right (delay): whatever happen in f (t) at some
instant t also happens in g(t) = f (t − T ), T seconds later at the instant
t + T (i.e., g(t + T ) = f (t)).

13
 5. Signal operations

(2) T < 0, the shift is to the left (advance)

• To time shift a signal by T, we replace t with t − T ( f (t − T ) is f (t)

time-shifted by T seconds).

• Example:

1.5
f(t)

0.5

−5 −4 −3 −2 −1 0 1 2 3 4 5
t (sec)
f ( t ).

2 2

1.5 1.5
f(t + 1)

f(t − 1)

1 1

0.5 0.5

0 0

−5 −4 −3 −2 −1 0 1 2 3 4 5 −5 −4 −3 −2 −1 0 1 2 3 4 5
t (sec) t (sec)

f (t + 1): f (t) advanced(left-shifted) by 1 second. f (t − 1): f (t) delayed(right-shifted) by

1 second.
Conventionally t is expressed in seconds except if specified otherwise.

• Do Example 1.3

• Do Exercise E1.5

5.2 Time inversion

Given a signal f (t), its time-inversed g(t) is given by g(t) = f (−t). f (−t) is
given by a reflection of f (t) about the vertical axis (mirror image of f (t) about
the vertical axis is f (−t) ): Whatever happens for f (t) at time instant t also
happens for g(t) = f (−t) at time instant −t (i.e., g(−t) = f (t)).

• To time invert a signal: replace t with −t.

14
 5.3. Time scaling

• Example:
1.2 1.2

1 1

0.8 0.8

0.6 0.6

f(−t)
f(t)

0.4 0.4

0.2 0.2

0 0

−0.2 −0.2
−5 −4 −3 −2 −1 0 1 2 3 4 5 −5 −4 −3 −2 −1 0 1 2 3 4 5
t (sec) t (sec)

f ( t ). f (−t) symmetric to f (t) with respect

to the vertical axis.
• Do Example 1.5

5.3 Time scaling

Given a signal f (t), its time-scaled g(t) is given by g(t) = f ( at) where a is a
constant. The time scaling is the compression/expansion of a signal in time.
(1) If | a| > 1, the scaling results in compression: Whatever happens in f (t)
at some instant t also happens to g(t) = f ( at) at the instant t/a (i.e.,
g( at ) = f (t))
(2) If | a| < 1, the scaling results in expansion
• Time scaling a signal by a factor a: t is replaced by at
• The origin t = 0 is the anchor point [the signal is fastened at t = 0].
It remains unchanged under scaling operation because at t = 0, f (t) =
f ( at) = f (0).
• For a < 0: Time scaling is implemented together with time inversion.
• Example:

1.5
f(t)

0.5

−5 −4 −3 −2 −1 0 1 2 3 4 5
t (sec)
f ( t ).

15
 5. Signal operations

2 2

1.5 1.5
f(2t)

f( 2t )
1 1

0.5 0.5

0 0

−5 −4 −3 −2 −1 0 1 2 3 4 5 −6 −4 −2 0 2 4 6
t (sec) t (sec)

f (2t): f (t) compressed in time by a factor of 2. f ( 2t ): f (t) expanded in time by a factor

2.
• Example 1.4
• Exercise E1.6

5.4 combined operations

Given f (t), its time-shifted, time-inversed and time-scaled g(t) is given by
g(t) = f ( at − b). The simultaneous use of more than one operation of the
three operations can be done by two methods: shift then scale [Time shift
by b and then Time scale the shifted signal f (t − b) by a (if a negative, also
implements time inversion)] and scale then shift [Time scale f (t) by a (if a
negative, also implements time inversion) and then Time shift f ( at) by ba (i.e.,
replace t by t − ba )].

• Example:
You are given f (t):

1.5
f(t)

0.5

−6 −4 −2 0 2 4 6
t (sec)
f ( t ).
Plot f (4 − 2t).
Solution:

1. f ( at − b) with a = −2 and b = −4.

16
 5.5. Amplitude transformations

2. Shift then scale

2

1.5
f(t)

0.5

−6 −4 −2 0 2 4 6
t (sec)
f ( t ).
2 2 2

1.5 1.5 1.5

f(−2t + 4)
f(2t + 4)
f(t + 4)

1 1 1

0.5 0.5 0.5

0 0 0

−8 −6 −4 −2 0 2 4 6 8 −8 −6 −4 −2 0 2 4 6 −8 8 −6 −4 −2 0 2 4 6 8
t (sec) t (sec) t (sec)

f ( t + 4). f (2t + 4). f (−2t + 4)

3. Scale then shift
2

1.5
f(t)

0.5

−6 −4 −2 0 2 4 6
t (sec)
f ( t ).
2 2 2

1.5 1.5 1.5

−4
f(−2(t − −2 )
f(−2t)
f(2t)

1 1 1

0.5 0.5 0.5

0 0 0

−8 −6 −4 −2 0 2 4 6 8 −8 −6 −4 −2 0 2 4 6 8 −8 −6 −4 −2 0 2 4 6 8
t (sec) t (sec) t (sec)

f (2t). f (−2t). f (−2(t − −4 )) = f (−2t + 4).

−2

5.5 Amplitude transformations

1. Amplitude shifting Given a signal f (t), its amplitude-shifted g(t) is
given by g(t) = f (t) + B.

17
 5. Signal operations

1.5
f(t)

0.5

−5 −4 −3 −2 −1 0 1 2 3 4 5
t (sec)
f ( t ).

3
1

0.8
2.5
0.6

0.4
2
0.2
f(t) + 1

f(t) − 1
0
1.5
−0.2

−0.4
1

−0.6

0.5 −0.8

−1

0
−5 −4 −3 −2 −1 0 1 2 3 4 5 −5 −4 −3 −2 −1 0 1 2 3 4 5
t (sec) t (sec)

f (t) + 1. f (t) − 1.

2. Amplitude reversal Given a signal f (t), its amplitude-reversed g(t) is

given by g(t) = − f (t): − f (t) is a mirror image of f (t) about the hori-
zontal axis.

2 0

1.5 −0.5
−f(t)
f(t)

1 −1

0.5 −1.5

0 −2

−5 −4 −3 −2 −1 0 1 2 3 4 5 −5 −4 −3 −2 −1 0 1 2 3 4 5
t (sec) t (sec)

f ( t ). − f ( t ).

3. Amplitude scaling Given a signal f (t), its amplitude-scaled g(t) is given

by g(t) = A f (t).

18
 2

1.5
f(t)

0.5

−5 −4 −3 −2 −1 0 1 2 3 4 5
t (sec)
f ( t ).
1.2
4

3.5 1

3
0.8

2.5
0.6
2f(t)

f(t)

2
2

0.4
1.5

1 0.2

0.5
0

0
−0.2
−5 −4 −3 −2 −1 0 1 2 3 4 5 −5 −4 −3 −2 −1 0 1 2 3 4 5
t (sec) t (sec)

f (t)
2 f (t): Amplified. 2 : Attenuated.

6 Signal models
So far we have mainly encountered 4 signal models: Sinusoid, Exponential,
Product of sinusoid and exponential, and Ramp signal. This section intro-
duces 3 models commonly used in area of signals and systems, namely the
step function, the impulse function and the exponential function with a gen-
eral complex exponent. These 3 signal models are used as basis to represent
other signals and they can be used to simplify different aspects of signals and
systems.

6.1 Unit step function u(t)

{
1 t≥0
u(t) =
0 t<0

• See Figure 1.14 (a)

(1) The unit step function u(t) is a causal signal and can be used to rep-
resent other causal signals or to segment other signals. A given ar-
bitrary signal f (t) possibly noncausal, is reduced to a causal signal

19
 6. Signal models

by multiplying it by u(t) ( f (t)u(t) is a causal signal); for example,

e−at u(t) is a causal signal.

1.2

2.5 1

0.8
2

0.6

f(t)u(t)
f(t)

1.5

0.4

1
0.2

0
0.5

−0.2
−1 −0.8 −0.6 −0.4 −0.2 0 0.2 0.4 0.6 0.8 1 −1 −0.8 −0.6 −0.4 −0.2 0 0.2 0.4 0.6 0.8 1
t (sec) t (sec)

e−t . e − t u ( t ).

(2) A unit step function is used to represent piecewise-defined signals:

Using unit step function, we can describe such signals by a single
expression that is valid for all t.

Example: Rectangular pulse (or function)

Express the following rectangular pulse in terms a combination of

delayed unit step function:

1.2

0.8

0.6
f(t)

0.4

0.2

−0.2
−1 0 1 2 3 4 5
t (sec)

Solution: f (t) = u(t − 2) − u(t − 4)

Express the following segmented ramp signal in terms of step func-

tions:

20
 6.1. Unit step function u(t)

3.5

2.5
f(t)

1.5

0.5

−1 0 1 2 3 4 5
t (sec)

Solution: f (t) = t(u(t − 2) − u(t − 4))

Express the following signal in terms of step functions

1.5
f(t)

0.5

−1 0 1 2 3 4 5 6 7 8
t (sec)

Solution:

f (t) = t(u(t) − u(t − 2)) + u(t − 2) − u(t − 4)

2 14
+ (− t + )(u(t − 4) − u(t − 7))
3 3
2
= tu(t) − tu(t − 2) + u(t − 2) − u(t − 4) − tu(t − 4)
3
14 2 14
+ u(t − 4) + tu(t − 7) − u(t − 7))
3 3 3
2 14
= tu(t) + (−t + 1)u(t − 2) + (−1 − t + )u(t − 4)
3 3
2 14
+ ( t − ) u ( t − 7)
3 3
2 11 2 14
= tu(t) + (−t + 1)u(t − 2) + (− t + )u(t − 4) + ( t − )u(t − 7)
3 3 3 3

• Do Example 1.6

• Do Example 1.7

21
6. Signal models

• Exercise E1.7

• Exercise E1.8
) (
t−T
• Rectangular pulse or function is expressed in general as Arect
a
[a is the width of the rectangular pulse while A is the height of the
rectangular pulse and T is the location of its centre].

( ) {
t−T A T − 2a ≤ t < T + a
2
Arect =
a 0 otherwise

Examples:

1. Unit rectangular pulse with a width τ:

1.2

0.8

0.6
f(t)

0.4

0.2

−0.2
−τ/2 0 τ/2
t (sec)

{
t 1 − τ2 ≤ t < τ
2
rect( ) =
τ 0 otherwise

2. One of the rectangular pulse that was given above u(t − 2) − u(t −
4) can also be represented as: u(t − 2) − u(t − 4) = rect( t−2 3 ).

• Homework:

1. f (t) = 2(t − 1)[u(t − 1) − u(t − 3)]

t
2. f (t) = t2 rect( )
3
3. f (t) = −u(t)
4. f (t) = u(−t)
5. f (t) = u(t)u(t − 1)
6. f (t) = et u(−t)

22
 6.2. Unit impulse function δ(t)

6.2 Unit impulse function δ(t)

• Also called Dirac delta function (or δ function) and has some analogous
properties to the Kronecker delta function encountered in the study of
discrete-time signals. The most important function in field of signals and
systems.

• First definition of a unit impulse function:

Definition 6.1 (classical definition by Paul Dirac).

{
δ(t) = 0 for t ̸= 0 (zero
∫ ∞ everywhere except at t = 0 where it is undefined)
−∞ δ (t )dt = 1 (unit impulse function has an area of 1)

• It is a theoretical signal and not physical; and there are different impulse
approximations (i.e., different intuitive descriptions of an impulse). Fol-
lowings are some approximating functions:
1/ε 1/ε 1/ε

2
f(t) = √ e−t /2²
u(t)

2
1 −t/²

² 2π
f(t)

f(t)

²→0
e

1
²
f(t) =

²→0 ²→0 ²→0

0 0 0

−ε/2 0 ε/2 −ε 0 ε −ε 0 ε 0
t (sec) t (sec) t (sec) t (sec)

Elevated, narrow Elevated, narrow Elevated, narrow Elevated, narrow

rectangular pulse of exponential pulse of triangular pulse of Gaussian pulse of
unit area. unit area. unit area. unit area.

• Plot of a unit impulse function:

2
f(t)

1.5

1
δ(t)

0.5

−2 −1.5 −1 −0.5 0 0.5 1 1.5 2

t (sec)

• Properties of an impulse function:

1. Given f (t) = Kδ(t):

Kδ(t) = 0 for all t ̸= 0 and the area of Kδ(t) = K: Impulse function
whose area is K
2. δ(t − T ): Impulse located at t = T

23
 6. Signal models

2 5

1.5 4.5

1 4
4δ(t − 3)
0.5 3.5
f(t)

f(t)
0 3

−0.5 2.5

−1 2

−1.5 1.5

−2
−2δ(t) 1

−2.5 0.5

−3 0

−2 −1.5 −1 −0.5 0 0.5 1 1.5 2 −4 −3 −2 −1 0 1 2 3 4

t (sec) t (sec)

−2δ(t): Its strength is 2 and amplitude re- 4δ(t − 3): Its strength is 4
versed. and located at 3.
The strength of an impulse refers to the area under the impulse;
and a unit impulse function δ(t) has a strength of 1
3. Multiplication of a function by an impulse:

ϕ(t)δ(t) = ϕ(0)δ(t) : given that ϕ(t) is continuous at t = 0

ϕ(t)δ(t − T ) = ϕ( T )δ(t − T ) : given that ϕ(t) is continuous at t = T
ϕ(t − T )δ(t) = ϕ(− T )δ(t) : given that ϕ(t) is continuous at t = − T

Example:

f(T)

δ(t − T )
f(t)

0
0 T

f (t)δ(t − T ) = f ( T )δ(t − T )
4. Sampling or sifting property of the unit impulse function

∫∞ ∫∞
ϕ(t)δ(t)dt = ϕ(0)δ(t)dt
−∞ −∞
∫∞
= ϕ (0) δ(t)dt
−∞
= ϕ (0)

given that ϕ(t) continuous at t = 0.

∫∞
It can also be shown that −∞ ϕ(t)δ(t − T )dt = ϕ( T ) where ϕ(t)

24
 6.2. Unit impulse function δ(t)

assumed to be continuous at t = T (T is the instant where the

impulse is located). The integral is referred to as sifting integral: It
sifts out the value of ϕ(t) at t = T.
Sampling/sifting property: The area under the product of a
function ϕ(t) with an impulse δ(t − T ) (or δ(t − T ) when T = 0)
is equal to the value of that function at the instant where the
unit impulse is located ϕ( T ).
Sampling property is a consequence of the first definition of im-
pulse function (the classical definition).

• Impulse function is neither an energy nor a power signal because it has

infinite energy and power. Proved by considering
( ) the impulse function
A t
as the limit of the rectangular pulse rect as ϵ → 0.
ϵ ϵ

• The first definition of the impulse function is not mathematically rigor-

ous

1. Impulse function does not define a unique function; e.g., δ(t) + δ̇(t)
also satisfy the first definition.
2. It is not a true mathematical function in ordinary sense (i.e., does
not specify a value of the dependent variable for each value of in-
dependent variable): It is undefined at t = 0.
3. It is defined as a generalized function rather than an ordinary func-
tion. A generalized function is defined by its effect on other func-
tions instead of its value at every instant of time. Impulse function
is defined in terms of its effect on a test function ϕ(t).

• Second definition of a unit impulse function:

Definition 6.2 (A generalized function approach). A unit impulse is a

function for which the area under its product with a function ϕ(t) is
equal to the value of the function ϕ(t) at the instant where the impulse is
located, assuming that ϕ(t) is continuous at the location of the impulse.

Consequently, any continuous function can be represented as a summa-

tion of weighted shifted impulses. In this second definition, the sam-
pling/sifting property defines the impulse function.

• Application of the generalized function definition of an impulse: u(t) is

discontinuous at t = 0 and dudt does not exist at t = 0 in ordinary sense;
however it does exist in the generalized sense and it is δ(t); i.e.,

du
= δ(t)
dt
(Proof from Equation 1.25 to Equation 1.27). Consequently:

25
6. Signal models

∫t {
0 t<0
δ(τ )dτ = = u(t)
1 t≥0
−∞

• Do Exercise 1.9
• Do Exercise 1.10
• Other examples:
1. (t3 + 3)δ(t) =
ω2 + 1
2. δ ( ω − 1) =
ω2
3. t2 δ(−t) =
∫0
4. [δ(t) − δ(t − 1) + δ(t + 1)]dt =
−4

6.3 Complex exponential function est

where s is a complex in general s = σ + jω

• In the chapter on The Background (Chapter 0) in the textbook, we re-

viewed on the one hand the exponetial function eα t and on the other
hand the polar representation re j θ of a complex number. And we are
able to deduce the following relations:

est = e(σ+ jω )t
= eσt (cos ωt + j sin ωt)

es̄t = eσt (cos ωt − j sin ωt)

where s̄ is conjugate of s. And the sum of the two relations gives:

1 st
eσt cos ωt = (e + es̄t )
2

• est is a generalization of the function e jωt (Euler’s formula). s is com-

plex frequency; |ω | is the frequency of oscillatoion of est and is known
as radian frequency; and σ is Neper frequency (attenuation/growth),
which gives the information about the rate of increase or decrease of the
amplitude of est .
• The function f (t) = Kest = Ke(σ+ jω )t encompasses a large class of func-
tions:

26
 1. σ = 0, ω = 0 =⇒ f (t) = K [A constant (a DC) signal] (see Figure
1.21 (a))
2. ω = 0 =⇒ f (t) = eσt [Monotonic increasing or decreasing expo-
nential] (see Figure 1.21 (a))
3. σ = 0 =⇒ Re{ f (t)} = cos ωt [A sinusoid] (see Figure 1.21 (b))
4. σ ̸= 0, ω ̸= 0 =⇒ Re{ f (t)} = eσt cos ωt [Exponentially varying
sinusoid] (see Figure 1.21 (c) and Figure 1.21 (d))

• Complex frequency s representation on a complex plane (s plane) (given

in Figure 1.22):
Signals whose complex frequency s, with respect to the s-plane, lie:

1. in the Left half plane: Exponential decaying signals

2. in the Right half plane: Exponential growing signals
3. on the jω-axis (imaginary axis): Signals of constant amplitude (i.e.,
sinusoids)
4. on the σ-axis (real axis): Exponential signals
5. on the origin: Constant signals

7 Even and odd functions

Even/Odd characteristics: Signal characteristics that describe the relationship
between signal values for positive time and signal values for negative time.

• f e (t) is an even function of t if

f e (−t) = f e (t)

An even function has the same value at the instants t and −t for all
values of t. f e (t) is symmetrical about the vertical axis (see Figure 1.23
(a)).

• f o (t) is an odd function of t if

f o (−t) = − f o (t)

For odd function, the value at the instant t is the negative of its value
at the instant −t. f o (t) is anti-symmetrical about the vertical axis (or
symmetric with respect to the origin) (see Figure 1.23 (b)).

7.1 Some properties of even/odd functions

• Multiplication:

1. Even function × Odd function = Odd function

2. Odd function × Odd function = Even function
3. Even function × Even function = Even function

27
7. Even and odd functions

• Area
∫a ∫a
f e (t)dt = 2 f e (t)dt
−a 0
∫a
f o (t)dt = 0
−a

7.2 Finding even and odd components of a signal

• Many signals are neither even or odd; and every signal f (t) can be ex-
pressed as a sum of even and odd components (any signal f (t) can be
decomposed into a sum of even f e (t) and odd f o (t) signals).

• Given that f e (t) = [ f (t) + f (−t)] is an even function of t and f o (t) =

[ f (t) − f (−t)] is an odd function of t, a function f (t) can be expressed
as

1 1
f (t) = [ f (t) + f (−t)] + [ f (t) − f (−t)]
2 2
i.e., a sum of an even f e (t) function and an odd function f o (t).

• Do Example 1.8

• Example: Decompose the function f (t) = e− at u(t) into its even and odd
parts.

• Express the following signal as a sum of an even and odd functions.

1.2

0.8

0.6
f(t)

0.4

0.2

−0.2
−2 −1.5 −1 −0.5 0 0.5 1 1.5 2
t (sec)

28