TRIBHUVAN UNIVERSITY

INSTITUTE OF ENGINEERING PULCH

OWK CAMPUS

Report on

Signal Analysis

Submitted by:
Ankush Jha

SUBMITTED TO:
DEPARTMENT OF COMPUTER
& ELECTRONICS ENGINEERING
LAB 1: TRANSFORMATION OF INDEPENDENT VARIABLES IN SIG
NAL ANALYSIS AND PRIMARY SINGLES.

1.1. OBJECTIVES

i.) To familiarize with the transformation of independent variables in signal

analysis (i.e. time shifting, time scaling and time inversion).

ii.) To utilize MATLAB for visualizing transformed signals and validating

theoretical results through simulations.

1.2. THEORY

A signal can be defined as a function of one or more independent variable whic

h carry information about the behaviour or nature of phenomena. For example,
human speech (sound wave), voltage in electrical circuits etc.

1.2.1. CLASSIFICATION OF ONE-DIMENSIONAL SIGNAL

a. Continuous-time signal: A signal x(t) is said to be a continuous time signal if

it is defined for all time t. A continuous signal is specified at every value of it
s independent variable. For example, the temperature of a room.

Figure1.1: Continuous-time signal

1
b. Discrete-time signal: A discrete time signal x[nt] has values specified only at
discrete points in time. As most practical signals are continuous signals, the
discrete signal is often obtained by sampling the continuous signal. For exa
mple, crime rate, budget of a country etc.

Figure1.2: Discrete-time signal

1.2.2. Some Standard Signals

Some primary signals that are very useful for signal analysis are listed below:
a. Sinusoidal signal: The signal varying with sine function of time or cosine fu
nction of time is sinusoidal signals. The general expression for sinusoidal sign
als is given below. x(t)=A sin(wt+∅) and Acos(wt+∅) are

Figure1.3: Continuous-time sinusoidal signal

b. Ramp signal: The signal that increases linearly with positive value of time an
d zero for negative value of time is ramp signal. Mathematically, continuou
s-time ramp signal is represented as,

0, 𝑡<0

2
 𝑟(𝑡) = {

𝑡, 𝑡≥0

And discrete-time ramp signal is represented as,

0, 𝑛<0

𝑟[𝑛] = {

𝑛, 𝑛≥0

Figure1.4(a): Continuous-time ramp Figure1.4(b): Discrete-time ramp sign

al signal

c. Unit- step signal: The signal that begins from the time instant ‘0’ and goes u
p to ‘+’with the constant amplitude of unity is unit-step signal. Mathematica
lly, continuous-time unit-step signal is represented as,

0, 𝑡<0

𝑢(𝑡) = {

1, 𝑡≥0

And discrete-time unit-step signal is represented as,

0, 𝑛<0

𝑢[𝑛] = {

3
 1, 𝑛≥0

Figure1.5(a): Continuous-time unit Figure1.5(b): Discrete-time unit-step st

ep signal signal

the amplitude oof unity is unit-impulse signal. It is represented by δ(t) in co

d. Unit-impulse signal: The signal that exists only at the time instant zero with

ntinuous time and δ[n] in discrete time respectively.

Figure1.6(a): Continuous-time unit Figure1.6(b): Discrete-time unitimpulse

impulse signal signal

e. Exponential signal: In continuous time, a real exponential signal can be repr

esented as,

x(t)=c𝑒𝑎𝑡 Where,
both c and a are real numbers.

Case-I: When a<0, signal will decay exponentially.

Case-II: When a>0, signal will grow exponentially.

4
 Figure 1.7(a): Growing exponential

Figure 1.7(b): Decaying exponential

1.2.3. TRANSFORMATION OF THE INDEPENDENT VARIABLES a.

Time Shift:

Figure 1.8: Discrete-time signals related by a time shift. In this figure n0>0, s
o that x[n−n0] is a delayed version of x[n]. (i.e., each point in x[n] occurs lat
er in x[n−n0]).

5
A simple and very important example of transforming the independent variabl
e of a signal is a time shift. A time shift in discrete time is illustrated in Figure 1.
8, in which we have two signals x[n] and x[n−n0] that are identical in shape, bu
t that are displaced or shifted relative to each other.

We will also encounter time shifts in continuous time, as illustrated in Figure 1.

9, in which x(t−t0) represents a delayed (if t0 is positive) or advanced (if t0 is ne
gative) version of x(t).

6
Figure 1.9: Continuous-time signals related by a time shift. In this figure t0<0, s
o that x(t−t0) is an advanced version of x(t) (i.e., each point in x(t) occurs at an
earlier time in x(t−t0)).

b. Time Reversal: A second basic transformation of the time axis is that of tim
e reversal. For example, as illustrated in Figure 1.10, the signal x[−n] is obta
ined from the signal x[n] by a reflection about n=0 (i.e., by reversing the sig
nal).

Figure 1.10: (a) A discrete-time signal x[n]; (b) its reflection x[−n] about n=0.

7
Similarly, as depicted in Figure 1.11, the signal x(−t) is obtained from the signal
x(t) by a reflection about t=0. Thus, if x(t) represents an audio tape recording, t
hen x(−t) is the same tape recording played backward.

Figure 1.11: (a) A continuous-time signal x(t); (b) its reflection x(−t) about t=0.

c. Time Scaling: Another transformation is that of time scaling.

8
In Figure 1.12 we have illustrated three signals, x(t), x(2t), and x(t/2), that are r
elated by linear scale changes in the independent variable. If we again think of
the example of x(t) as a tape recording, then x(2t) is that recording played at t
wice the speed, and x(t/2) is the recording played at half-speed.

9
Figure 1.12: Continuous-time signals related by time scaling.

It is often of interest to determine the effect of transforming the independent v

ariable of a given signal x(t) to obtain a signal of the form x(αt+β), where α and
β are given numbers.

Such a transformation of the independent variable preserves the shape of x(t),

except that the resulting signal may be linearly stretched if |α|<1, linearly com
pressed if |α|>1, reversed in time if α<0, and shifted in time if β is nonzero.
1.3. MATLAB PROGRAMS AND OUTPUTS 1.3.1.
FOR PRIMARY SIGNALS

a. Sinusoidal signals:

Continuous-time signal Discrete-time signal

Source code: t=-p Source code: n=-p
i:0.01:pi x=sin(t) pl i:1:pi x=sin(n) ste
ot(t,x,'r') xlabel('ti m(n,x,'r') xlabel('ti
me(t)') ylabel('x me(t)') ylabel('x
(t)') [n]')

title('sinusoidal signal') title('sinusoidal signal')

Output:

10
 Output:

b. Ramp signals:

Continuous-time signal Discrete-time signal

Source code: Source code:

c = input('Enter the intercept (c): c = input('Enter the intercept (c):

') m = input('Enter the slope (m): ') m = input('Enter the slope (m):
') t= -10:0.01:10 for i = 1:length(t) ') n = -10:1:10 for i = 1:length(n)
if t(i) < 0 y(i) = 0; else if n(i) < 0 y(i) = 0; else

y(i) = m * t(i) + c en y(i) = m * n(i) + c

d end end

end stem(n, y,'r') xlabel

('n') ylabel('x[n]') ti
plot(t, y,'r') xlabel tle('Ramp Signal')
('t') ylabel('x(t)') tit
le('Ramp Signal')

11
 Output: Output:
c=0 m=1 c=0 m=1

c. Unit-step signals

Continuous-time signal Discrete-time signal

Source code: t = - Source code: hol
5:0.01:5 u = doubl d on for(n=-10:1
e(t >= 0) plot(t, u) 0) if(n<0) s
xlabel('Time (t)') tem(n,0) else

ylabel('x(t)') stem(n,
1,'r') end end
title('Continuous-Time Unit Step xlabel('n') ylabel
('x[n]')
Signal')
title('Unit-step Signal') hold
off

Output:

12
 Output: Output:

d. Exponential signals
Continuous-time signal Discrete-time signal
Source code: Source code:

%y=ce^(aT) %y=ce^(aT)

%where c=amplitude ,scaling factor %where c=amplitude ,scaling factor

%a=nautre of exponential1 %a=nautre of exponential

1 n=-5:1:5 c=input('c=') a=i
t=-5:0.01:5 c=inp nput('a=') y=c*exp(a*n) st
ut('c=') a=input('a em(n,y) xlabel('n') ylabel
=') y=c*exp(a*t) p ('x[n]')
lot(t,y) xlabel('t')
ylabel('x(t)') title('exponential Signal')

title('exponential Signal')

13
 Output: Output:

I. c=1, a=1 I. c= 1, a=1

II. c= 1, a=-1

1.4. TRANSFORMATION ASSIGNMENT IN LAB

Q. Plot the following signal.

14
 a.
x(-t)

b. x (− 𝑡
+ 1)

d. x(t). 𝛿(𝑡)
c. x (2t + 1)

Answer:

a. x(-t) is the folded version of x(t) which will be,

15
b. To obtain x (− 𝑡
+ 1) ,first we perform time shifting of x(t) by 1 unit to obta
in 2

x(t+1), then we perform time inversion and then we will expand the signal by ½
unit as:

16
c. To obtain x (2t + 1) first we perform time shifting of x(t) by 1 unit to obtain x
(t+1), and then we will compress the signal by 2 unit as:

17
d. To obtain x(t). 𝛿(𝑡), we will get

1.5. DISCUSSION

In the first lab of Signal Analysis, we discussed about the primary signals and si
gnal transformation and visualized using MATLAB. We also discussed about the
precedence rule that we use while performing multiple transformation in a sig
nal.

18
1.6. CONCLUSION

In first lab of Signal Analysis, based on focused objectives to understand the pri
mary signals and transformation of a signal, the additional lab activities made u
s more confident towards the fulfilment of objectives.

Lab - 2 Convolution

1 Objective

1. Concept of Convolution
2. Convolution in Discrete and Continuous Cases
3. Convolution using MATLAB

2 Theory

2.1 Convolution

Convolution plays a crucial role in understanding LTI systems because it pro

vides a straightforward way to compute the system’s output for any given in
put. The key property of LTI systems is that their output can be entirely char
acterized by their impulse response h(t) or h[n].

By convolving the input signal with the impulse response, one can determin
e the system’s behavior and predict its response to other inputs. This proces
s leverages the superposition principle and the time-invariance property of t
he system, ensuring that the system’s response is consistent and predictable
regardless of the input signal’s characteristics. Convolution is fundamental i
n many areas of signal processing, control systems, and communications, as
it enables the analysis and design of complex systems with linear and time-i
nvariant properties.

2.2 Convolution in Continuous Domain

In continuous domain given input signal x(t) and an impulse response h(t) of
the system, the output can be constructed as:

19
 scanning axis

Figure 1: Convolution in Continuous Domain

2.3 Convolution in Discrete Domain

In continuous domain given input signal x(t) and an impulse response h[n] o
f the system, the output can be constructed as:

20
 Figure 2: Convolution in Discrete Domain

3 Convolution in MATLAB

In this lab we performed convolution using MATLAB for continuous and disc
rete cases for finite and infinite duration signals and impulse response.

3.1 Discrete Time Finite Impulse Response

Consider input x[n] and h[n] as following:

x[n] = {1,3,2,1,1} h[n] = {1,2,−1,1}

Source Code
x=[1,-1,0,1]

h=[1,0,0,1]

y=conv(x,h)

subplot(311)

stem(x)

subplot(312)

stem(h)

subplot(313)

21
stem(y)

OUTPUT

Numerical Result:

x = 1 -1 0 1

h=1 0 0 1

y = 1 -1 0 2 -1 0 1

Fig. Graphic
al output

3.2 Discrete Time Infinite Impulse Response

Consider input x[n] and h[n] as following:

22
 x[n] = u[n] h[n] = (1/2)n

The Source Code in MATLAB is as follows:

n=0:10

x=0.5.^n

h=ones(1,11)

y=conv(x,h)

subplot(311)

stem(x)

subplot(312)

stem(h)

subplot(313)

stem(y)

OUTPUT

Numerical Output:

x = Columns 1 through 10

1.0000 0.5000 0.2500 0.1250 0.0625 0.0312 0.0156 0.0078 0.0039

0.0020

Column 11

0.0010

h= 1 1 1 1 1 1 1 1 1 1 1

y = Columns 1 through 10

23
1.0000 1.5000 1.7500 1.8750 1.9375 1.9688 1.9844 1.9922 1.9961
1.9980

Columns 11 through 20

1.9990 0.9990 0.4990 0.2490 0.1240 0.0615 0.0303 0.0146 0.0068

0.0029

Column 21

0.0010

Fig. Graphical Output

3.3 Continuous Time Infinite Impulse Response

Consider input x[n] and h[n] as following:

x(t) = e−atu(t) h(t) = u(t)

Source Code

24
hold on

t=0:0.5:5

x=exp(-0.5*t)

h=ones(1,11)

y=conv(x,h)

subplot(311)

plot(x)

subplot(312)

plot(h)

subplot(313)

plot(y)

hold off

OUTPUT

Numerical Output:

t = Columns 1 through 10

0 0.5000 1.0000 1.5000 2.0000 2.5000 3.0000 3.5000 4.0000 4.5

000

Column 11

5.0000

x = Columns 1 through 10

25
1.0000 0.7788 0.6065 0.4724 0.3679 0.2865 0.2231 0.1738 0.1353
0.1054

Column 11

0.0821

h=1 1 1 1 1 1 1 1 1 1 1

y = Columns 1 through 10

1.0000 1.7788 2.3853 2.8577 3.2256 3.5121 3.7352 3.9090 4.0443

4.1497

Columns 11 through 20

4.2318 3.2318 2.4530 1.8465 1.3741 1.0062 0.7197 0.4966 0.3228

0.1875

Column 21

0.0821

Fig. Graphical Output

4 Discussion

26
In this lab, we used MATLAB to explore the concept of convolution in both di
screte and continuous cases. By defining and plotting discrete signals, we pe
rformed convolution using MATLAB’s built-in functions, allowing us to obser
ve the resulting output signals clearly. This exercise demonstrated the proce
ss of manipulating and analyzing discrete-time signals through convolution
operations.

We also extended our understanding to continuous signals by defining an

d plotting continuous-time functions. Using the integral definition of convolu
tion, we computed and visualized the output of convolving an exponential de
cay signal with a unit step function. This lab session provided a comprehensi
ve understanding of convolution and implementation in MATLAB in perform
ing signal processing tasks efficiently.

5 Conclusion

This lab concluded with the implementation of the concept of convolution in

both discrete and continuous cases and implementation in MATLAB, to visul
aize the output for different functions and impulse responses. The visualizati
ons and computations performed during this lab provided valuable insights
into the behavior of Linear Time-Invariant (LTI) systems and the impact of c
onvolution on input signals.

LAB 3: FOURIER TRANSFORM AND FOURIER SERIES

1 Objective

1. Concept of Fourier Series and Transform

2. Implementation in MATLAB
3. Fast Fourier Transform

27
2 Theory
2.1 Fourier Series

The Fourier Series is a mathematical tool used to represent a periodic functi

on as a sum of sine and cosine functions, or more compactly, as a sum of com
plex exponentials. This series decomposes a complex periodic signal into its
fundamental frequency components, making it easier to analyze and underst
and the behavior of the signal in the frequency domain. Each term in the Fou
rier Series corresponds to a specific harmonic of the fundamental frequency,
with its own amplitude and phase.

For a periodic function in continuous domain, the Fourier Series is an infinit

e series:

where, Cn is the spectral coefficient and T is the time period of fundamental h

armonic.

For a periodic function in discrete domain, the Fourier Series is a finite serie
s:

armonic. And ⟨N⟩ means bounded by N, which means N successive terms of

where, Cn is the spectral coefficient and T is the time period of fundamental h

k.

28
 2.2 Fourier Transform

The Fourier Transform is a powerful mathematical tool used to convert a tim

e-domain signal into its frequencydomain representation. Unlike the Fourier
Series, which applies to periodic signals, the Fourier Transform applies to bo
th periodic and aperiodic signals, making it extremely versatile in analyzing
a wide range of signals and systems.

For Continuous Case, the Fourier Transform of signal f(t) is given by:

For Discrete Case, the Fourier Transform of signal x[n] is given by:

X(ejω) = X x[n]e−j
ωn n=− ∞

2.3 Spectral Analysis

The analysis of spectral coefficients of periodic signals (both continuous and

discrete) or in general analysis of Fourier Transform or signals is called Spec
tral Analysis. Since, the Fourier Transform is complex in nature, we analyze t
he magnitude part separtely to analyze the strength of each harmonic compo
nent in the signal and phase plot to anaylze the phase shift or time alignment
of each frequency component.

Different type of Signals admit different magnitude and phase plot but their
nature is common which can be summarized as followed:
Magnitud Phase
e
CTFS Discrete Discrete

Aperiodic Aperiodic

Symmetric Anti-Symm

29
 etric
Discrete Discrete

DTF Periodic Periodic

S
Symmetric Anti-Symm
etric
Continuous Continuous

CTF Aperiodic Aperiodic

T
Symmetric Anti-Symm
etric
Discrete Continuous
Periodic
DTF
Aperiodic
T Asymmetri
Symmetric c

Table 1: Summary of the Nature of Spectrum

2.4 Fast Fourier Transform

The Fast Fourier Transform (FFT) is an efficient algorithm used to compute t

he Discrete Fourier Transform (DFT) and its inverse. The FFT significantly r
educes the computational complexity of the DFT, making it practical for anal
yzing and processing large datasets and real-time signals. The DFT converts
a discrete-time signal into its frequency-domain representation by decompo
sing the signal into a sum of sinusoidal components, each with a specific amp
litude and phase.

The FFT achieves this by leveraging the symmetry and periodicity propert
ies of the sinusoidal basis functions, which allows it to reduce the number of
required calculations. While a direct computation of the DFT involves O(N2)
operations for a signal of length N, the FFT reduces this to O(N logN) operati
ons. This efficiency gain is particularly important in applications such as digi
tal signal processing, communications, audio and image processing, and scie
ntific computing.

There are several variants of the FFT algorithm, with the most well-know
n being the Cooley-Tukey algorithm. This algorithm recursively breaks down

30
a DFT of any composite size N = N1 × N2 into smaller DFTs of sizes N1 and N2, c
ombining the results to obtain the final transform. By recursively applying th
is decomposition, the algorithm efficiently handles large inputs and produce
s the frequency-domain representation of the signal.

The FFT is a cornerstone of modern signal processing and has enabled nu

merous advancements in various fields by making the analysis and manipula
tion of frequency-domain data both feasible

3 Implementation in MATLAB
3.1 Fourier Transform of Discrete Aperiodic Signal

Consider the signal x[n] = {...,0,0,0,1,2,3,0,0,...} The source code to plot its Spe
ctral Analysis i.e. Phase and Magnitude Plot is as below:
x=[0,1,2,3]

y=fft(x,4)

subplot(211)

stem(real(y))

subplot(212)

stem(imag(y))

OUPUT

Numerical Output:

x=0 1 2 3

y = 6.0000 + 0.0000i -2.0000 + 2.0000i -2.0000 + 0.0000i -2.0000 - 2.0000

i

31
 Fig. Spectral Analysis of discrete aperiodic signal

3.2 Fourier Transform of Continuous Aperiodic Signal

Consider the signal below:

Source Code
x=[0,1,2,3]

y=fft(x,4)

subplot(211)

plot(real(y))

subplot(212)

plot(imag(y))

32
 Figure 2: Spectral Analysis of Aperiodic Continuous Signal

From the above spectrum, it can be observed that the magnitude plot is co
ntinuous, aperodic and symmetric. Also the phase plot ideally must be a rect
angular wave oscillating between 0, −π and π but due to resolution and preci
sion issue, it could not be a perfect rectangular wave. The phase plot thus obt
ained is continuous, aperiodic and anti-symmetric.

3.3 Fourier Transform of Discrete Periodic Signal

Consider the discrete periodic signal below having the principle part given b
elow and period N = 20:

1 −2 ≤ n ≤ 2

x(t) =

0 Elsewhere

Source Code

33
x=[1,1.5,0.7,2.3,6,1,0,0]

y=fft(x,6)

subplot(211)

stem(real(y))

subplot(212)

stem(imag(y))

OUTPUT

Numerical Output:

x = 1.0000 1.5000 0.7000 2.3000 6.0000 1.0000 0 0

y = Columns 1 through 5

12.5000 + 0.0000i -3.4000 + 4.1569i -1.3000 - 5.0229i 2.9000 + 0.0000i -1.300

0 + 5.0229i

Column 6

-3.4000 - 4.1569i

34
 Fig. Spectral Analysis of Discrete Periodic Signal

3.4 Fourier Transform of Continuous Periodic Signal

Consider the continuous periodic signal below having the principle part give
n below and period N = 20:

x=[1,1.5,0.7,2.3,6,1,
0,0]

y=fft(x,6)

subplot(211)

plot(real(y))

subplot(212)

35
plot(imag(y))

OUTPUT

Numerical Output:

x = 1.0000 1.5000 0.7000 2.3000 6.0000 1.0000 0 0

y = Columns 1 through 5

12.5000 + 0.0000i -3.4000 + 4.1569i -1.3000 - 5.0229i 2.9000 + 0.0000i -1.300

0 + 5.0229i

Column 6

-3.4000 - 4.1569i

Figure 4: Spectral Analysis of Continuous Periodic Signal

36
4 Discussion

In this lab we understood the concept of spectral analysis on various signals

to extract the information about the presence of harmonic in a signal and the
ir analysis using a powerful technique called Fourier Transforms. The conce
pt was implemented using MATLAB and the poweful and fast algorithm of F
ast Fourier Transform using fft command available in MATLAB tool box. We e
xtracted the magnitude and phase information by the application of MATLA
B and plotted magnitude and phase plots. The magnitude plots gives the info
rmation about the strength of particular frequency in a signal, while the phas
e plot gives the time shift of those frequency in the signal.

We experimented with signal which were Discrete: Aperiodic and Periodic a

nd Continuous: Aperiodic and Periodic. Thus, we verified the nature of magn
itude and phase plot which was summarized in the Table 1.

5 Conclusion

This lab concluded with the implementation of Fourier Transform in MATLA

B using fft function that uses Fast Fourier Transform to efficiently convert a s
ignal into frequency domain for different classes of signals discrete and conti
nuous, periodic and aperiodic. And then their magnitude and phase plot wer
e plotted to analyze the contribution of different range frequencies in a signa
l and their time shifts.

LAB 4: FREQUENCY RESPONSE ANALYSIS

OBJECTIVES

a. To understand the phase response and magnitude response of a system

b. To understand the concept of plotting phase and magnitude response of giv
en transfer function using MATLAB.

37
THEORY

The frequency response of a system is a fundamental concept in signal processi

ng and control systems. It describes how a system responds to sinusoidal input
s of different frequencies.

For digital system, the transfer function of system is expressed as,

H(z)=𝑏10++𝑎𝑏11𝑧𝑧−−11++𝑎𝑏22𝑧𝑧−−22++⋯⋯++𝑏𝑎𝑚𝑛𝑧𝑧−−
𝑛𝑚

The zeros of system can be obtained as,

𝑏0 + 𝑏1𝑧−1 + 𝑏2𝑧−2 + ⋯ + 𝑏𝑚𝑧−𝑚 = 0

The zeros determine the frequencies at which the system's output is zero (or m
inimized) for a given input.

The poles of system can be obtained as,

1 + 𝑎1𝑧−1 + 𝑎2𝑧−2 + ⋯ + 𝑎𝑛𝑧−𝑛 = 0

The poles determine the stability and transient response of the system.

. MAGNITUDE AND PHASE RESPONSE

If poles or zeros of system is expressed in polar form as,

𝑧 = 𝑟∡𝜃

Where r = radius of poles and zeros

𝜃 = angle of poles and zeros

38
MATLAB PROGRAMS AND OUTPUTS

39
Output:

40
:

DISCUSSION

In the fourth lab of Signal Analysis, we discussed about the frequency response
of the digital system. Here, we were able to visualize how the system modified
the magnitude and phase of a signal at different frequencies using MATLAB. Th
e command plot (abs ()) and plot (angle ()) were used in MATLAB to plot the m
agnitude and phase response of the given system respectively.

We also discussed about the application of frequency response in real life. It ha

s been widely used in audio engineering, control system,telecommunication.

CONCLUSION

In fourth lab of Signal Analysis, we learnt and understand the frequency respo
nse of system using MATLAB.

41