DIGITAL SIGNAL PROCESSING(DSP)

OUTLINES:
1. BASIC ELEMENTS OF DSP
2. Z TRANSFORM (ZT)
3. DISCRETE FOURIER TRANSFORMS (DFT)
4. FAST FOURIER TRANSFORMS (FFT)
5. IIR (Infinite Impulse Response) FILTER DESIGN
6. FIR(Finite Impulse Response) FILTER DESIGN
7. FINITE WORD LENGTH EFFECTS
8. DIGITAL SIGNAL PROCESSORS
03/02/23 Dr Ngenzi Alexander 1
 Basic elements of Digital Signal Processing:
 DSP definition
 Revision on signals and systems
 Limitations of Analog Signal Processing
 Pros & cons of Digital Signal Processing
 Main DSP Applications
 Basic elements of DSP and their function
 Sampling theorem
 Discrete-time signal representation

03/02/23 Dr Ngenzi Alexander 2

 Common Discrete Time Signals
 Discrete Time Signal Expression
 Manipulations of Discrete Time Signal
 Classification of Signals
 Discrete Time Systems
 Classification of Systems
 Analysis of Linear time invariant systems

03/02/23 Dr Ngenzi Alexander 3

 DSP Definition
DSP: any manipulation on a variable carrying
information that can be conveyed
displayed
manipulated

It is concerned with the digital representation of

signals by a sequence of numbers or symbols and
the use of digital processors to analyze, modify,
or extract information from signals.
03/02/23 Dr Ngenzi Alexander 4
 Digital Signal Processing
It includes all transformations and manipulations
of signals such as:
filtering,
compressing and decompressing,
encoding and decoding,
analysis and representation,
speech-to-text conversion, etc.

03/02/23 Dr Ngenzi Alexander 5

 Definitions
 Signal: a physical quantity that varies with
time, space or any other independent
variables or invariables which convey
information.
 System: physical device that performs an
operation on the signal. Ex: A filter used to
reduce noise in the information bearing signal
is called a system
 Processing: operation performed by the
system on the signal.
03/02/23 Dr Ngenzi Alexander 6
 Examples of signals
Electrical signals
 Voltages and currents in a circuit
Acoustic signals
 Acoustic pressure (sound) over time
Mechanical signals
 Velocity of a car over time, forces,
pressures
Video signals
 Intensity level of a pixel (camera, video)
over time
03/02/23 Dr Ngenzi Alexander 7
 Signal Types

03/02/23 Dr Ngenzi Alexander 8

 Signal Types representations
Continuous time –
Continuous amplitude

Continuous time –
Discrete amplitude

Discrete time –
Continuous amplitude

Discrete time –
Discrete amplitude
03/02/23 Dr Ngenzi Alexander 9
 Why signal processing?
Why signals? Signals carry information.
Why processing? To obtain signals in a more
desirable form, a clean form.
Signals are corrupted by noise, they need to
be filtered.
Signal processing can be:
Analog, Digital or Mixed

03/02/23 Dr Ngenzi Alexander 10

 Reason for processing
• The specific reason for processing a digital
signal may be, for example, to remove
interference or noise from the signal, to
obtain the spectrum of the data, or to
transform the signal into a more suitable
form.

03/02/23 Dr Ngenzi Alexander 11

 Digital Signal Processing
Digital Signal Processing uses unique type of data
i.e. signal, for processing signals. It refers to the
science of analyzing time-varying physical
process.
A signal refers to any continuous function of one
or more variables such as time, space, frequency,
etc. e.g.
Voltage across a resistor
Velocity of a vehicle
Light intensity of an image
Temperature or pressure inside a system
03/02/23 Dr Ngenzi Alexander 12
 Signal Processing
There are two categories of signal processing:
1. Analog Signal Processing
Used to describe a waveform that is
continuous in time and can take a continuous
range of amplitude values. It will be more
correct to say continuous signal processing.
2. Digital Signal Processing
Used to describe a discrete-time-signal, not a
continuous waveform but discrete-time
signal quantities with one amplitude value of
signal at discrete instants in time.
03/02/23 Dr Ngenzi Alexander 13
 First Signal Processing Problem
World War II brought us Radio Detection And

Ranging (RADAR)
 One of the first signal processing applications
 It was implemented in analog
 Encounter lots of problem, aging, stability, sensitivity…

03/02/23 Dr Ngenzi Alexander 14

Information Representation Using
Analog Signals
Information can be represented using analog signals

Analog signals cannot be manipulated easily

Analog signals must be digitized for computer

processing i.e. changed in binary form for computer
processing

03/02/23 Dr Ngenzi Alexander 15

 Digital Signal Processing
Advantages Limitations
•High noise immunity& • Bandwidth limitations
so limited speed of
& high accuracy operation of AD
• Upgradability & Flexibility converters and digital
• Stability & Repeatability signal processors
• Higher performance and • System complexity:
lower cost usage of ADC or DAC
• Great flexibility of its • DSP systems more
programmability expensive
• Perfect Reproducibility • More power
• Special Applications consumption more
• Saving of bandwidth components
03/02/23 Dr Ngenzi Alexander 16
 Advantages of DSP over ASP (Analog
Signal Processing)
1. Noise-resistance and high accuracy
2. Great flexibility: DSP operations can be
changed by changing the program in digital
programmable system.
3. Stability and repeatability: easily duplicated
and independent of temperature, aging and
other external parameters.
4. Simplicity: easy to build any digital system
5. Easy upgradations: because of use of software
03/02/23 Dr Ngenzi Alexander 17
6. Compatibility: all applications need standard
hardware. Thus operations of DSP system are
mainly dependent on software.
7. Remote processing: Easy information storage
on magnetic media without loss of quality of
reproduction of signal and easily transported,
digital signals can be processed off line.
8. In some cases, cheaper, digital hardware
cheaper and digital circuits can be reproduced
easily in large quantities or by time-sharing of
given processor among a number of signals.

03/02/23 Dr Ngenzi Alexander 18

 DSP Applications
Have replaced analog circuitry in TV sets and
telephones & commonplace in such devices as
mobile phones, multimedia computers, video
recorders, CD players, hard disc drive controllers
and modems.

03/02/23 Dr Ngenzi Alexander 19

03/02/23 Dr Ngenzi Alexander 20
03/02/23 Dr Ngenzi Alexander 21
 What Is DSP?

03/02/23 Dr Ngenzi Alexander 22

 Examples of DSP System

03/02/23 Dr Ngenzi Alexander 23

 Temperature Control System
 

03/02/23 Dr Ngenzi Alexander 24

 DSP System Overview
• Receives a signal, processes it, and outputs
result
• Several DSPs can be linked together in a serial
chain
• Can be implemented in software or hardware

03/02/23 Dr Ngenzi Alexander 25

 Key DSP Operations

Convolution

Correlation (Cross & Auto)

Filtering

Discrete transform

03/02/23 Dr Ngenzi Alexander 26

 Basic Elements of DSP

03/02/23 Dr Ngenzi Alexander 27

 Function of basic elements of DSP
1. Analog I/P Signal generated from some
transducer or from some communication system
or biomedical signal like ECG Electrocardiogram
(heart signals) or EEG Electroencephalogram(brain
signals).
2. Low pass filter (Antialiasing filter) removes the
high frequency noise contain in input signal.

3. Analog to digital converter (ADC): used to convert

analog signal into digital form, because digital
signal processor accepts the signal which is digital.
 
03/02/23 Dr Ngenzi Alexander 28
4. Digital Signal Processor : It processes input signal
digitally. It can be a digital computer or a
microprocessor programmed to perform the
desired operations on the input signal or also a
hardwired digital processor.
5. Digital to analog converter (DAC): the output of
digital signal processor is digital in nature. But
the required final output is analog in nature. So
to convert digital signal into analog signal.
6. Output Filter signal is analog that means it is
continuous signal. But it may contain high
frequency components. Such high frequency
components are unwanted. To remove these
components output filter is used.
03/02/23 Dr Ngenzi Alexander 29
 Analog to Digital Conversion
Converts a continuous, analog signal into a
discrete, digital signal
Analog signal is sampled at a regular interval
Sampled values are stored as a data stream

03/02/23 Dr Ngenzi Alexander 30

 Analog to Digital Conversion
1 0 1 1 0 1 0 0

A to D Converters, Digital Signal Processors

(DSP)
03/02/23 Dr Ngenzi Alexander 31
 Data Transmission Using Analog
Technology

Computer Modem

Digital: 0s and 1s Analog

Digital-to-Analog Modulation
Modem from mo ( dulator ) dem ( odulator )] is an
electronic device that makes possible the transmission of
data to or from a computer via telephone or other
communication lines.
03/02/23 Dr Ngenzi Alexander 32
 Digital Information Processing

Data
Audio Digitized
Digital
and
Image Transmission
Encoded
Video

03/02/23 Dr Ngenzi Alexander 33

From Continuous to Discrete: Sampling

03/02/23 Dr Ngenzi Alexander 34

 Sampling theorem

where : x(t)= sinωt or x(t)= sin2πfot;

fo, Frequency in Hz and t time in seconds
If the highest frequency component in a signal is fmax,
then the signal should be sampled at the rate of at
least 2 fmax . Ex. If fmax is 4kHz, then to preserve or
capture all the information in the signal, sample it at
8Khz or more
03/02/23 Dr Ngenzi Alexander 35
 Sampling Theorem
Nyquist - Shannon sampling theorem honors
Harry Nyquist and Claude Shannon
 A signal must be sampled at a rate at least 2x
the frequency of the maximum frequency
component (Nyquist frequency) or else aliasing
will occur [Nyquist (1928), Shannon (1949)]
Harry Nyquist was an important contributor to
communication theory
Aliasing is the inability to accurately reconstruct
a signal from sampled data.
Low-pass analog filter (anti-aliasing filter) applied
before sampling.
03/02/23 Dr Ngenzi Alexander 36
 The discrete-time waveform x(nTs)

If x(t)= sinωt or x(t)= sin2πfot; Ts : sampling time

period and n: number of samples
Then x(nTs)= sin2πfo nTs
Fs is the sampling frequency = 1/Ts
x(nTs) is a representation of x(t) that is sampled
once every T seconds using an ADC.
03/02/23 Dr Ngenzi Alexander 37
 Sampling

03/02/23 Dr Ngenzi Alexander 38

 Discrete-time signal representation
1.Functional representation:

2.Sequence representation:

3. Tabular representation

4.Graphic representation:
03/02/23 Dr Ngenzi Alexander 39
 Common Discrete Time Signals
Impulse Signal
 Unit impulse (or unit sample):
0, n  0
 [ n]  
1, n  0

Delayed unit sample or shifted unit impulse

03/02/23 Dr Ngenzi Alexander 40

 Common Discrete Time Signals
Unit Step Signal
Unit step sequence u[n]
1

0, n  0
u[n]  
1, n  0
n
0

Delayed or shifted unit step sequence.

03/02/23 Dr Ngenzi Alexander 41

 Common Discrete Time Signals
Ramp Signal

03/02/23 Dr Ngenzi Alexander 42

 Common Discrete Time Signals

03/02/23 Dr Ngenzi Alexander 43

 Common Discrete Time Signals

03/02/23 Dr Ngenzi Alexander 44

 Common Discrete Time Signals

03/02/23 Dr Ngenzi Alexander 45

 Sinusoidal Sequence

Period is 2π/ω only when 2π/ω is an integer.

03/02/23 Dr Ngenzi Alexander 46
Discrete Time (DT) Signal Expression
•x(n) be a discrete time signal
•δ(n-k) be a delayed impulse
signal
•If we multiply x(n) with the
delayed impulse δ(n-k),
•then the product is non-zero
only at n = k and zero for all
other values of n. Also, at n =
k, the value of the product
signal is the kth sample of
signal x(k).
03/02/23
•x(n) δ(n-k) = x(k)
Dr Ngenzi Alexander 47
 DT Signal Expression
Each multiplication of x(n) with δ(n-k) at
same delay k picks out the signal value x(k)
of the signal x(n) at n=k. If we repeat this
from
-∞ < k < ∞, and sum all the product
sequences, the result will be a sequence of
that is equal to x(n).

x[ n]   x[k ] [n  k ]
k  
Any discrete-time signal can be expressed
as a summation of impulses.
03/02/23 Dr Ngenzi Alexander 48
 Example
x(n) δ((n-(-3)) = x(-3)
x(-3) = x(-3) δ((n-(-3))
x(n) δ((n-(-2)) = x(-2)
x(-2) = x(-2) δ((n-(-2))
x(n) δ(n-(-1)) = x(-1)
x(-1) =x(-1) δ((n-(-1))
x(n) δ(n) = x(0)
x(0) = x(0) δ((0)
x(n) δ(n-1) = x(1)
x(1) = x(1) δ((n-1))
x(n) δ(n-2) = x(2)
x(2) = x(2) δ((n-2
x(n) δ(n-3) = x(3)
Thus we can say that x(n) is the product of the
samples and the delayed impulses:
x(n)= …… + x(-3)+ x(-2)+ x(-1)+ x(0)+ x(1)+ x(2)+ …
+x(…)+…..
= ….. x(-3) δ((n-(-3))+ x(-2) δ((n-(-2))+ x(-1) δ((n-(-1))+
x(0) δ((0)+ x(1) δ((n-1))+Drx(2)
03/02/23
δ((n-2)+ ……
Ngenzi Alexander 49
:

Sum of Weighted Shifted Impulses

Any sequence, x[n], can be expressed

as a sum of scaled, delayed impulses
03/02/23 Dr Ngenzi Alexander 50
 Relationship between impulse
and
unit step signal

The impulse signal is the first difference of the

unit step signal

 [n]  u[n]  u[n  1]

The unit step signal is the running sum of the
impulse signal 
u[n]    [n  k ]
n
u[n]    [ m]
m   k 0
03/02/23 Dr Ngenzi Alexander 51
 Manipulations of Discrete Time Signal
 Shift in time : x[n] it is the operation that
gives us either delayed version x [n-2]or an
advanced version x [n +2]
Folding or reflection : from x[n] it is the
operation that gives us the signal x[−n], replace n
by -n of x(n) about n = 0
 Addition, multiplication and scaling of
sequences :
Amplitude scaling: y [n] = A x [n]
Sum of 2 signals : y [n] = x1 [n] + x2 [n]
Product of 2 signals: y [n] = x1 [n] . x2 [n]
03/02/23 Dr Ngenzi Alexander 52
Manipulations of Discrete Time Signal

03/02/23 Dr Ngenzi Alexander 53

Manipulations of Discrete Time Signal

03/02/23 Dr Ngenzi Alexander 54

Manipulations of Discrete Time Signal

03/02/23 Dr Ngenzi Alexander 55

Manipulations of Discrete Time Signal

03/02/23 Dr Ngenzi Alexander 56

Manipulations of Discrete Time Signal

03/02/23 Dr Ngenzi Alexander 57

Manipulations of Discrete Time Signal

03/02/23 Dr Ngenzi Alexander 58

Manipulations of Discrete Time Signal

03/02/23 Dr Ngenzi Alexander 59

 Classification of Signals
Energy & Power Signals
For a discrete-time signal x[n]
Normalized energy content E of x[n]

Normalized average power P of x[n]

03/02/23 Dr Ngenzi Alexander 60

 Classification of Signals
Energy & Power Signals
1. A signal x[n] is said to be an energy signal
if and only if , and so .

2. A signal x[n] is said to be a power signal if

and only if , and so

3. Signals that satisfy neither property are

referred as neither energy signals nor
power signals
03/02/23 Dr Ngenzi Alexander 61
 Classification of Signals
Periodic vs. Aperiodic Signal
A discrete-time signal x[n] is said periodic if
x[n] = x[n + N] for all values of n, x[n]) is
periodic with period N because it repeats
itself
after a fixed period N, otherwise it is called
aperiodic or non periodic.

03/02/23 Dr Ngenzi Alexander 62

 Classification of Signals

03/02/23 Dr Ngenzi Alexander 63

 Classification of Signals

03/02/23 Dr Ngenzi Alexander 64

 Continuous-Time System
x(t) → y(t)

03/02/23 Dr Ngenzi Alexander 65

 Discrete-Time System
x[n] → y[n]

03/02/23 Dr Ngenzi Alexander 66

 Continuous-Time &Discrete-Time Systems
A system operates on a signal (the input) to modify,
transform or re-express it in another form (the output)
which may be more desirable.
Example of system: The input voltage E(t) produces an
output voltage Vo(t) .

Continuous-time input signals are applied and result in

continuous-time output signals.
Discrete-time input signals are applied and result in discrete
03/02/23 Dr Ngenzi Alexander 67
time output signals.
 Discrete-Time Systems
• A system is a physical device that performs an
operation on the signal.
• Example: A filter used to reduce noise in the
information bearing signal is called a system.
• In contexts from signal processing to
communications, a system: a process in which
input signals are transformed by the system or
cause the system to respond in some way,
resulting in outputs signals.
Discrete-time system:
x[n] → y[n]
03/02/23 Dr Ngenzi Alexander 68
 Classification of Systems

Systems may be classified into:

1.Time-invariant and time-variant systems

2.Linear and non-linear systems

3.Causal and non-causal systems

03/02/23 Dr Ngenzi Alexander 69

Time-invariant and Time-Variant System
• A system is called time- or shift-invariant if a time
shift (delay or advance) in the input signal causes
the same time shift in the output signal.

• A system that is not time-invariant is time varying.

Example: y[n] = x[M n], - ∞ < n < ∞
• By considering the response y1[n] to the input
x1[n] defined as x1[n] = x[n - n0]:
y1[n] = x1[M n] = x[M n - n0]
But y[n - n0] = x[M (n - n0)] ≠ y1[n] unless of course
M = 1.
03/02/23 Dr Ngenzi Alexander 70
 Linear Systems
•A linear system exhibits the additivity property:

•It also must satisfy the homogeneity or scaling

property:

•These can be combined into the property of

superposition:

03/02/23 Dr Ngenzi Alexander 71

 Non-linear system
A non-linear system is one that is NOT linear
(i.e. does not obey the principle of
superposition). At least one of the above
properties (additivity or scaling) does not
hold
Examples of non-linear systems:
y = x2
T[x1 + x2 ]= (x1 + x2)2 = x12 + x22 + 2 x1x2

≠ T[ x1] + T [ x2 ] = x12 + x22

03/02/23 Dr Ngenzi Alexander 72
 Linear Time-Invariant Systems

If the system is linear and also time

Invariant or shift invariant, then it is

called a linear time invariant(LTI)or

a linear shift invariant (LSI) system.

03/02/23 Dr Ngenzi Alexander 73

 Causal and Non causal Systems
 A system is causal if the output at any time
depends only the present and/or past values of the
input, not on its future values
example:

A system is called non causal if it depends on its

future values
Examples of non causal systems are

03/02/23 Dr Ngenzi Alexander 74

 INDICATIVE RESOURCES
1. John G Proakis, Dimtris G Manolakis, “Digital
Signal Processing Principles, Algorithms and
Application”, PHI, 4th Edition, 2006,
2. Emmanuel C. Ifeacher, Barrie W. Jervis,
“Digital Signal Processing A Practical Approach,
1993
3. B Venkataramani & M Bhaskar, Digital Signal
Processors Architecture, Programming and
Applications, 2nd Edition, 2011, 004.16 VEN
2011.
4. M.H.Hayes, “Digital Signal Processing”,
Schaum’s Outlines, McGraw-Hill, 1999.
03/02/23 Dr Ngenzi Alexander 75