University of Dar es Salaam

ES 330: Telecommunication I

11 December 2023 1
 University of Dar es Salaam
Outline
2. Communication signals
2.1 Introduction

2.2 Time and frequency domain characteristics

2.3 Analysis of signals by Fourier series

2.4 Analysis of signals by Fourier integrals

2.5 Linear time invariant transmission networks

2.6 Energy and power signals

11 December 2023 2
 University of Dar es Salaam
2.1 Introduction The frequency domain representation then gives the
frequency characteristic of the models and the
In the analysis and synthesis of communication signals used.
systems, mathematical models are predominantly
used. Tools used for analyze and characterize communication
systems and signals are:
In these models, the characteristics systems
employed are treated using linear time-invariant ➢ Fourier series
methods. ➢ Fourier transform methods

In order to visualize the signals in the systems, both

Signals can also be characterized by the amount of
time and frequency domain representations are
energy they carry as well as the power they contain.
used.

11 December 2023 3
 University of Dar es Salaam
2.2 Time and frequency domain characteristics
There are several signal classifications.
A signal is defined as the time history of a These includes:
quantity - usually a voltage or current.
➢ Continuous time and discrete time signals
At any instant, the frequency components of the
➢ Analog and digital signals
signals can be observed.
➢ Periodic and aperiodic signals
𝑓(𝑡) = the instantaneous signal
➢ Deterministic and probabilistic signals

𝐹 𝜔 = the frequency component signal ➢ Energy and power signals

f(t) having frequency 𝜔

11 December 2023 4
 University of Dar es Salaam
2.2.1 Continuous time and discrete time signals Discrete time signal is a signal that is specified only at
at descrete points of 𝑡 = 𝑛𝑇.
Continuous time signal is a signal that is specified for
every value of time 𝑡.

Figure. Continuous time signal Figure. Discrete time signal

Examples: Audio and video Example: quarterly gross domestic product, monthly
sales of a company and stock market day average etc

11 December 2023 5
 University of Dar es Salaam
2.2.2 Analog and digital signals
A digital signal is a signal whose amplitude can take
A signal whose amplitude can take any value in a
on only a finite number of values.
continuous range is Analog signal

(a) (b) (c) (d)

Fig. analog and Fig. analog and Fig. digital and Fig. digital and
continuous discrete continuous discrete

11 December 2023 6
 University of Dar es Salaam
2.2.3 Periodic and aperiodic signals A signal 𝑓 𝑡 is aperiodic if it is not periodic

A signal f 𝑡 is said to be periodic if there exist Example: 𝑓 𝑡 = 𝐴𝑒 𝑗(𝜔0𝑡+𝜃) , −∞ ≤ 𝑡 ≤ ∞

a positive constant 𝑇0 such that:

𝑓 𝑡 = 𝑓 𝑡 + 𝑇0 −∞≤𝑡 ≤∞

Fig. Periodic signal Fig. Aperiodic signal

11 December 2023 7
 University of Dar es Salaam
2.2.4 Deterministic and random signals

Deterministic signal is a signal whose physical description

is known completely, either in mathematical form or a
graphical form.

A signal that is known only in terms of probabilistic description,

such as mean value, mean square value and distribution rather
than its full mathematical or graphical description is a random
signal

Noise and messages that are used to convey information are

examples of random signals

11 December 2023 8
 University of Dar es Salaam
2.2 Basic signals
Sinusoidal Signals The period of the sinuoid is
A sinusoidal signal is of the form
T = f1 2π
=
ω
x (t) = cos(ωt + θ).

where the radian frequency is ω, which has the with the units of seconds.
units of radians/s. The phase or phase angle of the signal is θ, given in
Also very commonly written as radians.

x (t) = A cos(2πft + θ).

where f is the frequency in Hertz.

We will often refer to ω as the frequency, but it
must be kept in mind that it is really the radian
frequency, and the frequency is actually f .
11 December 2023 9
 University of Dar es Salaam
2.2 Basic signals…

11 December 2023 10
 University of Dar es Salaam
2.2 Basic signals…

11 December 2023 11
 University of Dar es Salaam
2.2 Basic signals…

Uses of unit step

11 December 2023 12
 University of Dar es Salaam
2.2 Basic signals…

11 December 2023 13
 University of Dar es Salaam
2.2 Basic signals…

11 December 2023 14
 University of Dar es Salaam
2.2 Basic signals…

11 December 2023 15
 University of Dar es Salaam
2.2 Basic signals…
Shifting property

11 December 2023 16
 University of Dar es Salaam
2.2 Basic signals…

11 December 2023 17
 University of Dar es Salaam
2.2 Basic signals…

11 December 2023 18
 University of Dar es Salaam
2.3 Time and frequency domain characteristics…

The time representation of the real component of a signal and the frequency representation of the
real component are given in Figures below.
Amplitude

A/2 A/2

t w
-w0 w0

-
-A
𝜃
T

Figure. A plot of x 𝑡 = Acos(𝜔0 𝑡 + 𝜃) Figure. Frequency domain representation of

x 𝑡 = Acos(𝜔0 𝑡 + 𝜃)
11 December 2023 19
 University of Dar es Salaam
2.4 Analysis of signal by Fourier series where

The Fourier analysis of signals is either done by the

1 𝑡0 +𝑇
Fourier series or the Fourier integral. 𝑋𝑛 = න 𝑥(𝑡)𝑒 −𝑗𝑛𝜔0𝑡 𝑑𝑡
𝑇 𝑡0

In Fourier series, a portion of signal x(𝑡) in the interval and

𝑇 can be represented by Fourier series as follows:
2𝜋
 𝑇= , 𝑛 = −∞, … , −1,0,1, … ∞
x(t) =  X n e jnw0t , t 0  t  t 0 + T,
𝜔0

-
If the signal is periodic then the representation
applies to all t

11 December 2023 20
 University of Dar es Salaam
2.4 Analysis of signal by Fourier series

Consider a periodic signal 𝑥(𝑡), with fundamental Home work:

frequency 2𝜋 given as
3 1. Extend in long form
𝑥 𝑡 = ෍ 𝑋𝑛 𝑒 𝑗𝑛2𝜋𝑡
𝑛=−3
2. Apply Euler's principle to get the time domain signal

where
𝑋0 = 1
1
𝑋1 = 𝑋−1 = ,
4
1
𝑋2 = 𝑋−2 = ,
2
1
𝑋3 = 𝑋−3 = .
3

11 December 2023 21
 University of Dar es Salaam
Xn

2.4 Analysis of signal by Fourier series 1

1/2 1/2
Example: -2 -1 0 1 2 n
-2w0 -w0 0 w0 2w0 w

As an example, consider the signal

x(t) = 1 + cos( w0 t) + sin(2 w0 t) - cos(2 w0 t) Figure: The real frequency components of a

Fourier series signal.
By using Euler's the signal can be represented as

1 1 j j 1 1
x(t) = 1+ e jw0t + e- jw0t - e j2w0t + e- j2w0t - e j2w0t - e- j2w0t .
2 2 2 2 2 2

The frequency components of the real part are given in

the following Figure
Figure: The imaginary frequency components of
11 December 2023 a Fourier series signal. 22
 University of Dar es Salaam
2.4 Analysis of signal by Fourier series… where
a0 2a 2b
X0= = A0 , An = n ,bn = n
If the signal is periodic then the representation applies T T T

to all t
a n - j bn , a n + j bn
X n= X -n =
cos(nw t)+ j sin(nw t) = A +  A cos(n w
  T T
x(t) =  X n 0 0 0 n 0 t) + B n sin(n w 0 t)
T/2 T/2
- n =1 2 2
An = 
T -T/2
x(t) cos(n w0 t)dt, Bn =  x(t)sin(n w0 t)dt.
T -T/2

The signal is called the Trigonometric Fourier series. T/2 T/2

an = 
-T/2
x(t) cos(n w0 t)dt, bn =  x(t)sin(n w t)dt.
-T/2
0

with
Another representation of Fourier series is as follows:
T/2

 cos(n w
-T/2
0 t) cos(m w 0 t)dt = 0, n  m,
 
x(t) =  a n e jnw0t = 0 +  a n cos(n w 0 t) + bn sin(n w 0 t)
a 2
T/2
- T T n =1 T
 cos(n w
-T/2
0 t) sin(m w 0 t)dt = 0 n, m.

11 December 2023 23
 University of Dar es Salaam
2.4 Analysis of signal by Fourier series… The Fourier series representation of signals is used for
periodic signals.
Another representation of the Fourier series,
called the complex exponential representation.  
Example:  A,| t |

P (t) =  2
1 
C
2
jnw 0 t
x(t) = n e , 0,otherwise
T -

T/2

Cn =
-T/2
 x(t) e- jnw0 t dt.
A

This gives
t
T 2 2 j n - /2  /2
C n = an - j bn = a + b e = ( An - j Bn )
n n -T
T
2
 bn 
where  n = tan -1  -  ,
 an 

and *
C n = an + j bn complex conjugate of Cn. Figure. Rectangular pulse of amplitude 𝐴

11 December 2023 24
 University of Dar es Salaam
2.4 Analysis of signal by Fourier series…
/2 /2

C=  Ae- jwt
dt = - e  = 2A sin( w/2)= A sin( w/2)
A - jwt  A

- /2
jw -/2 w w/2

sin(u)
= A
u

0 u
where u = w/2.
-3 -2 - 0  2 3
Noting that
lim sin(u)
=1
u →0 u

sin(u) 
=0 m = 1,2,3,.. Figure. Plot of the C function for the
u u=m rectangular pulse.
11 December 2023 25
 University of Dar es Salaam
2.4 Analysis of signal by Fourier series…
Example: Consider the following function In this case we obtain w = n w,0 where w0 = 2/T.


x(t) =  P (t - nT), T >
n=- 
2

A sin( u n )
Therefore: Cn = sin(n w0 /2)= A
n w0 /2 un
A

u n = n w0 /2.
-T -  /2  /2
T
where

Figure: Plot of an infinite train of rectangular pulses.

11 December 2023 26
 University of Dar es Salaam
2.3 Analysis of signal by Fourier series…

The frequency spectrum contains values ofC n scaled by a factor sin(u)/u and separated by w = 2/T

A
n=-1 n=1

-3 -2 -  2 3 u
0

-8 /T -4 /T -2 /T 2 /T 4 /T 6 /T w

Figure: Frequency spectrum of an infinite train of rectangular pulses.

11 December 2023 27
 University of Dar es Salaam
where w0 = 2/T
2.4 Analysis of signal by Fourier integral

The analysis of periodic signals is performed by lim lim

Defining F( w ) = Cn and f(t) = x(t)
the use of Fourier series. T → T →

However, the signals encountered are non- and noting lim

n w0 = w
periodic and the tool best suited for their analysis T →
is the Fourier Integral.


To develop the Fourier Integral, consider the We obtain

F( w ) = 
-
f(t) e- jwt dt.
Fourier series complex representation  
1
 T/2 x(t) =  e =
C n jnw0t
 Cn e jnw ot
w
x(t) =  C n e jnw0t , C n =  x(t) e- jnw0t dt
Like wise
- T 2 -

- T -T/2
where
w = w0 = 2/T
11 December 2023 28
 University of Dar es Salaam
2.4 Analysis of signal by Fourier integral… 

The infinite sum above can be represented by the

F( w ) =
-
 f(t) e-wt dt

Riemann integral

1
 F( w ) e dw
jwt
 f(t) =
1 2
f(t) =
2 
-
F( w ) ewt dw . -

The (Fourier Transform) relationship between the two

The Fourier transform and inverse Fourier integrals is denoted by
transform are given by the following pair of
integral, respectively
f(t)  F( w ).

11 December 2023 29
 University of Dar es Salaam
2.4 Analysis of signal by Fourier integral… lim
PT (t) =  (t).
T →0
An important non-periodic signal is the unit
impulse function denoted by  (t) and defined as
For the sinc function (inverse transform), the first zero
 crossing is given by
 f(t) (t)dt = f(0)
- 
w 1 T =  => w 1 =
where 𝑓(𝑡) is a test function which is continuous T
at 𝑡 = 0 lim
w1 = 
lim sin( wT)
and we note T →0 , therefore =1
T → 0 wT
Example: Consider the rectangular pulse PT(t)
having a unite area. implying that the sinc function spreads out to attain unit
amplitude throughout the frequency spectrum as the
sin( wT) sin( wT) width T of the rectangular pulse approaches zero
PT (t)  =
wT wT  (t)  1.
11 December 2023 30
 University of Dar es Salaam
2.4 Analysis of signal by Fourier integral…

Consider the signal

Figure: Fourier transform of the signal 𝑥(𝑡)

The fourier transform above is complex valued, to
plot it as a function of 𝜔, we express 𝑋(𝑗𝜔) in
terms of its magnitude and phase Homework: Given the signal

Plot the signal in time domain, find its fourier

transform and draw 𝑋(𝑗𝜔)
11 December 2023 31
 University of Dar es Salaam
2.4 Analysis of signal by Fourier integral… 6. FrequencyTranslation : f(t) e jw0t  F( w - w0 )

1 1
Fourier Transforms have a number of properties 7. Modulation : f(t) cos( w0 t)  F( w - w0 ) + F( w + w0 )
2 2
which can be proved mathematically and are given
as theorems. d
n
8. Differentiation : n
f(t)  (j w )n F( w )
dt
1. Linearity: a1 f 1 (t) + a2 f 2 (t)  a1 F 1 ( w ) + a2 F 2 ( w )
t
1 1
9. Integration :  f(  )d  F( w ) + F(0) ( w )
jw
2.Time Delay : f(t - t 0 )  F( w ) e-jw t 0 -
2

1 10.Convolution : f 1 (t) * f 2 (t)  F 1 ( w ) F 2 ( w )

3. Scaling : f(at)  F( w/a)
|a|
1
4.Time Reversal: f(-t)  F(-w ) = F ( w ) * 11. Multiplication : f 1 (t) • f 2 (t)  F 1 ( w )* F 2 ( w )
2

5. Duality : F(t)  f( w )

11 December 2023 32
 University of Dar es Salaam
2.4 Analysis of signal by Fourier integral…

Example: Given the signal signal.

1 5 5
𝑥 𝑡 = 𝑥1 𝑡 − + 𝑥2 (𝑡 − )
2 2 2

a. Draw the signal 𝑥(𝑡)

b. Find its fourier transform
Figure: The signal 𝑥1 𝑎𝑛𝑑 𝑥2

11 December 2023 33
 University of Dar es Salaam
2.4 Analysis of signal by Fourier integral… From the differentiation theorem
d
The properties of the transform pairs can be used
to solve Fourier transform problems for the
u(t)  jwF( w )
analysis of signals with much greater flexibility.
dt
d
Example: consider the unit step (a non-periodic
and u(t) =  (t),  (t)  1
dt
signal) defined by:
d
u(t) =  (t)  jwF( w ) = 1
1, t  0
then
u(t) =  dt
1
0, t < 0 therefore u(t)  F( w ) = .
jw
Let
u(t)  F( w ). Therefore, by using known transforms of simpler signals,
the theorems can be used to obtain transforms of more
11 December 2023 complicated signals. 34
 University of Dar es Salaam
2.5 Linear time invariant transmission networks Such a network is characterised by

A transmission network (or system) can be 1. The impulse response h(t) such that g(t) = h(t)
represented by the following block diagram when f(t) =  (t)

2. 𝐻(𝜔) is the Fourier transform of ℎ(𝑡)


f(t) h(t) g(t)
3. g(t)= f(t)* h(t)=  f( )h(t -  )d  G( w ), = F( w )H( w )
-

F(w) H(w) G(w)

4. H( w ) = | H( w ) | e j ( w )
Fig. Block diagram of a linear time-invariant
network. The network scales the input signals by | H( w ) | and
modifies their phases by e  w j ( )

11 December 2023 35
 University of Dar es Salaam
2.5 Linear time invariant transmission networks…
H(w)

If the network is also linear and time-invariant,

then it is further characterized by
w

1. a1 f 1(t) + a2 f 2 (t) → a1 g 1,(t) + a2 g 2 (t) -W W

Fig. Plot of the transfer function of an ideal low-
2. f(t -  ) →
. g(t -  ) pass filter.

The impulse response of the filter is given by

A typical transmission network is the ideal filter
 W
characterized by its transfer function 1 1
- H( w ) e dw = 2  dw
jwt jwt
h(t)= 1 • e
2
 1, | w | W -W
H( w ) = PW ( w ) = 
0, | w |> W =
2jt
e -W =
1 jwt W sin(Wt)
t
.
11 December 2023 36
 University of Dar es Salaam
2.5 Linear time invariant transmission networks…
Transmission networks are used in the processing of
The response of such a filter to a rectangular pulse
telecommunication signals.
PT(t) can be found as follows
2T sin( wT)
g(t) = PT (t) * h(t)  G( w ) = • PW (w) The processing follows one or more of the transform
wT theorems with changes in the amplitude-frequency
F(w) 2T
characteristics and phase modifications.
w
-3 -2 - 0  2 3

H(w) 1

w
-3/2 G(w) 2T 3/2

Fig. Response of an ideal low-pass filter to a

rectangular pulse.
11 December 2023 37
.
University of Dar es Salaam
2.5 Energy and power signals In such cases

   1  
Signals can be classified as Energy signals and or
  f * (t)  F( w )e dw  dt
jwt
E= f(t) f * (t)dt =
 2
Power signals.  -

- -

The energy in a signal f(t) is given by

lim T 
=
1
2



-
F( w ) 


-
f * (t)e jwt dt dw
T → 
E= | f(t)| dt =  | f(t)| dt
2 2

-T -
*
where | f(t)| is the instantaneous power.
2 1

 
 1


 w  w  w ( w )dw.
- jwt
 
*
and hence E = F( ) f(t) e dt d = F( ) F
2 - -  2 -
The signal f(t) is defied as an energy signal iff
 
0 < E < ,=> P = 0 E =  | f(t)| dt =
1
 w dw
2 2
Therefore | F( )|
-
2 -

11 December 2023 38
 University of Dar es Salaam
2.5 Energy and power signals… The power in a signal is given by

The above relationship, known as Rayleigh's T

lim
1

2
Energy Theorem (or Perseval's Theorem for Fourier P= | f(t)| dt.
transforms), gives the energy spectral density T →  2T -T
defined by
The signal f(t) is defined as a power signal iff
S( w ) = | F( w )|2
0 < P < ,=> E = 

from which T
lim

2
In such cases P= |< f(t) > | dt
 T →  -T
1
E= 
2 -
S( w )dw
f(t)
< f(t) > =
where 2T
11 December 2023 39
 University of Dar es Salaam
2.5 Energy and power signals…

Like in energy signals, this gives

  
1 1
P =  |< f(t) > | dt =
2
- |< F( w ) > | dw = 2
2
 S( w )dw
-
2 -

Where S( w ) is the power spectral density.

Thus the average power of a periodic signal is the sum of
the component powers in the Fourier series.
By applying Perseval's theorem to periodic signals,
we obtain

T/2 
Spectral densities are quite useful in determining where
1
P =  | f(t)|2 dt =  | X n|2 .
the power in signals is concentrated in the frequency
spectrum.
T -T/2 -
11 December 2023 40