Lecture 1 Introduction

by

Dr. Anirban Dasgupta

Assistant Professor, Dept. of Electronics and Electrical Engineering
Associated Faculty, Dept. of Design
Associated Faculty, Centre for Intelligent Cyber Physical Systems
IIT Guwahati
 What is a Signal?
• A signal is a manifestation of some physical quantity which usually varies with respect to
some independent variable, e.g. space or time, and contains some information.
• It is mathematically represented as a function of the independent variable.

Let us record the ambient

temperature and plot it as a
function of time

The signal is represented as

a function of the
independent variable, time
 Examples of Signals
amplitude

amplitude
time time

Speech signal Electrocardiography (ECG) signal

locations
amplitude

time

time
Seismic signal Electroencephalography (EEG) signal
 Signal Acquisition

Varying physical Signal in

quantity electrical form
Sensor Sampler
Analog signal Discrete-time
signal
Digital signal Quantized signal
Digital
Encoder Quantizer
Hardware

For storage, processing or transmission

 Signal Classification

• Continuous vs. Discrete Signals

• Deterministic vs. Stochastic Signals
• Even vs. Odd Signals
• Periodic vs. Aperiodic Signals
• Power vs. Energy Signals
 Continuous vs. Discrete Signals
• Continuous Signals:
• A continuous signal is defined for every point in time.
• It can take any value in a given range, and there are no breaks or
interruptions in the signal.
• Example: Analog audio signals
• Discrete Signals:
• A discrete signal is only defined at specific time intervals.
• It takes values at these discrete time steps, and between those intervals,
the signal is not defined.
• Example: A sequence of temperature readings taken every minute
 What is a Discrete-Time Signal?

A discrete-time signal is represented as , where is the signal

amplitude at sample index , where .

𝑥 [ 𝑛 ] =[3 ,2 , 5 , 6 , 0 , 7 , 6 ]
zero index,

A discrete-time signal can be obtained by

• sampling an analog signal , or
• even generated from a digital hardware.
 Discretization of Analog Signal

Sampler

𝑥 ( 𝑡 )=sin ( 2 𝜋 .2 𝑡 ) =sin ( 4 𝜋 𝑡 ) 𝑥 ( 𝑛𝑇 𝑠 )=sin ( 2 𝜋 .2𝑛𝑇 𝑠 )=sin ( 4 𝜋 𝑛𝑇 𝑠 ) , 𝑛∈ ℤ

Sampling
Discretization of Analog Signal

Discretizing
the time axis

𝑥 [ 𝑛 ] =sin ( 4 𝜋 𝑛 𝑇 𝑠 ) ,𝑛 ∈ ℤ 𝑥 (𝑛 𝑇 𝑠) 𝑥 [𝑛 ]
Digitalization of Analog Signal

discretizing
the
amplitude
values into
levels

usually levels

Quantization
 Important Discrete-Time Signals
Unit impulse Unit step function

{
𝛿 [ 𝑛 ] = 0 ,𝑛 ≠ 0
1 ,𝑛=0 {
𝑢 [ 𝑛 ] = 0 , 𝑛< 0
1,𝑛≥ 0

𝛿 [ 𝑛 ] =𝑢 [ 𝑛 ] −𝑢 [𝑛 −1] 𝑢 [ 𝑛 ] = ∑ 𝛿 [ 𝑛− 𝑚 ]
𝑚=0
 Important Discrete-Time Signals
Unit ramp signal Unit exponential decay

{ 𝑥 [ 𝑛 ] =𝑎 𝑢 [ 𝑛 ] ,|𝑎|<1
𝑛
𝑟 [ 𝑛 ] = 0 ,𝑛< 0 𝑟 [ 𝑛 ] =𝑛𝑢[𝑛]
𝑛 ,𝑛 ≥ 0
Important Discrete-Time Signals
discrete-time complex exponential
𝑗 ( 𝜔0 𝑛 + 𝜙 )
𝑥 [ 𝑛 ] =𝑒

Condition for periodicity

𝑚
𝜔 0 =2 𝜋 , 𝑚 ,𝑛 ∈ ℤ
𝑛

𝜔 0 =0.1 , 𝜙=0
 Deterministic and Stochastic Discrete-Time Signals
A signal is said to be deterministic if there is no A signal is said to be stochastic if the signal
uncertainty with respect to its value at any sample amplitude at each sample index comes with an
index. uncertainty.

𝑗 ( 𝜔0 𝑛 + 𝜙 )
𝑥 [ 𝑛 ] =𝑒 , 𝜔 0= 0.1 , 𝜙=0
 Even and Odd Discrete-Time Signals
Even signal is a signal which satisfies the relation Odd signal is a signal which satisfies the relation

𝑥 [ 𝑛 ] =𝑥 [ −𝑛 ] , ∀ 𝑛∈(− ∞, ∞) 𝑥 [ 𝑛 ] =− 𝑥 [ − 𝑛 ] , ∀ 𝑛∈(− ∞ , ∞)
Even signal is symmetric about the amplitude-axis Odd signal is anti-symmetric about the amplitude-axis

1 1
𝑥 [ 𝑛 ] = ( 𝑥 [ 𝑛 ] + 𝑥 [ −𝑛 ] ) + ( 𝑥 [ 𝑛 ] − 𝑥 [ −𝑛 ] )
2 2
 Periodic and Aperiodic Discrete-Time Signals
A discrete-time sequence is periodic, if the sequence The minimum positive value of is called its period,
repeats after samples. measured in samples
𝑥 [ 𝑛 ] =𝑥 [ 𝑛+𝑁 ] , 𝑁 ∈ ℤ , ∀ 𝑛∈(− ∞ , ∞) Any signal which does not obey the relation is
aperiodic
 Power and Energy Discrete-Time Signals
lim 1 𝑁

∑ |𝑥 [ 𝑛 ]|
𝑁 →∞ 2
signal energy signal power 𝑃=
2 𝑁 +1 𝑛=− 𝑁
If is finite, the signal is called an energy signal. If is finite, the signal is called a power signal.

The power for an energy signal is zero and the energy for a power signal is infinite.
 What is Frequency?
• Frequency refers to the number of cycles a sinusoidal signal completes per unit of time.
• Typically measured in Hertz (Hz), where 1 Hz equals one cycle per second.
• Frequency content and fundamental frequency are different

Both signals have same

fundamental frequency of 50 Hz

But the frequency contents are

different
 What is Frequency?
• Fundamental Frequency
• is the lowest positive frequency of a
periodic signal, often considered
the "primary" frequency
• Frequency content
• refers to the distribution of different
frequencies that make up a signal

All signals have same fundamental

frequency of 50 Hz, but different
frequency content
 Introduction to Fourier Analysis

How to know the frequency content of a signal?

• Frequency content refers to the different frequency components that
constitute a signal.
• By performing Fourier analysis on a signal, one can identify the
fundamental frequency and its harmonics and determine their relative
magnitudes.
• Periodic signals have discrete frequencies, multiples of fundamental
• Aperiodic signals have continuous frequencies
 Introduction to Fourier Analysis
What is Fourier Analysis?
• A mathematical technique to decompose signals into their frequency components.
• Used to analyze and process signals in the frequency domain.
• Fourier analysis works on the principle that any periodic signal can be represented
as a sum of sine and cosine waves at different frequencies.
• Fourier analysis are of two types
• Fourier Series: Used for periodic signals. It expresses a periodic function as
the sum of sines and cosines.
• Fourier Transform: Used for non-periodic signals, representing the signal in
the frequency domain.
 Line Spectra Vs Continuous Spectra

Line Spectra
• Line spectra refer to signals that consist of discrete frequency components.
• These components appear as sharp peaks or "lines" in the frequency
domain
• Signal is periodic.
Continuous Spectra
• Continuous spectra refer to signals whose frequency components are spread
over a continuous range
• Signals are aperiodic in nature
Line Spectra Vs Continuous Spectra
 Zero Frequency Vs Infinite Frequency

Zero Frequency:
• Zero frequency refers to the Direct Current (DC) component of a signal,
which corresponds to the average of the signal.
• A zero frequency component in a signal means that the component does not
change over time.
Infinite Frequency:
• Infinite frequency corresponds to the highest possible frequency.
• This concept arises in the context of idealized signals, as in the case of
perfectly sharp transitions or impulses.
Zero Frequency Vs Infinite Frequency

Zero Frequency

Infinite Frequency
 Negative and Complex Frequencies
• Frequencies can be both positive and negative
• Negative frequencies correspond to counterclockwise rotations
• A pure sinusoidal wave (e.g., sin(2𝜋
𝑓𝑡)) can be represented as a sum of two
complex exponentials:
• one with a positive frequency 𝑓 and
• the other with a negative frequency −𝑓
• Negative frequency represents the time-reversal of the signal
• Complex frequencies are frequencies that have both real and imaginary parts.
• Complex frequencies involve both oscillations and decay/growth.
Positive and Negative Components
 Concept of Phase
• Can two different signals have the same spectra?
• If yes, how to distinguish?
 Magnitude and Phase Spectra

• Spectra are of two types – magnitude and phase

• Magnitude Spectrum shows how much of each frequency component is present in
the signal.
• It quantifies the amplitude of each frequency component in the signal.
• The magnitude of the Fourier Transform of a signal is given by

• Phase Spectrum provides information about how each frequency component is

shifted in time relative to a reference.
• It shows the phase shift of each frequency component.
• The phase of the Fourier Transform of a signal is given by
 Sampling Frequency

Sampling frequency (​) refers to the number of samples taken per second, also measured
in Hz.
Why Sampling Frequency is Important?
• When converting a continuous-time signal to a discrete-time signal, need to pick
samples the continuous signal at a specific rate.
• Nyquist-Shannon Sampling Theorem states that the sampling frequency ​must be at
least twice the highest frequency component or bandwidth of the signal to
accurately reconstruct a signal from its samples and avoid aliasing.
• This minimum required sampling frequency is called the Nyquist rate: where is the
maximum frequency of the signal.
• Example: If a signal has frequency components up to 500 Hz, you need to sample at a
rate at least 1000 Hz (twice the highest frequency) to accurately reconstruct the signal
without aliasing.
Aliasing In Time Domain
Aliasing In Frequency Domain
• Folding frequency
is half sampling rate
• Represents the
maximum frequency
that can be accurately
represented in a
sampled signal
• Frequencies above the
folding frequency are
"folded" or aliased into
the lower frequency
range.
Aliasing In Frequency Domain
 Normalized Frequency
Normalized frequency is a concept used to express the frequency of a signal relative to
its sampling frequency.

​is the frequency of the signal in Hz.

​is the sampling frequency (in Hz).
is the normalized frequency, which is a dimensionless quantity.
The normalized frequency has a range of -0.5 to 0.5 for real-valued signals.
Frequencies greater than 1 are subject to aliasing and are folded back into the [-0.5,0.5]
range.
Example: If you have a signal with a frequency f=200 Hz and you are sampling it at
=1000 Hz, the normalized frequency is

This means the signal has a frequency that is 20% of the sampling rate.
 Digital Frequency
Digital Angular Frequency refers to the angular frequency of a discrete-time signal,
expressed in radians per sample

​is the digital angular frequency of the signal

​is the sampling frequency (in Hz).
is the normalized frequency
The digital angular frequency has a range of - to for real-valued signals.
Frequencies greater than are subject to aliasing and are folded back into the [- , ] range.
Example: Let's say we have a signal with a frequency of 100 Hz and the sampling
frequency ​ is 1000 Hz. The digital angular frequency