BERN 3763-DIGITAL SIGNAL PROCESSING

Chapter 1: Introduction to Digital Signal Processing

March 2025
Learning Outcomes
By the end of the chapter, you are expected to be able to:
• Explain what is signal, system & digital signal processing
(DSP).
• Recognise the importance and benefits of DSP.
• Distinguish between digital and analog signals.
What is signal?
• Variable that carries or contains some kind of information
• The information can be conveyed, displayed or manipulated
• Examples: speech signal, electrical signal (voltage & current),
temperature, pressure, etc.
• y(t), v(t) are dependant variables
• t is independent variable
What is signal?
• Can vary with time, frequency, space, distance, etc
• Example: Speech signal can be represented by amplitudes, frequencies
and phase (i.e. sum of several sinusoids)

 Ai (t) sin[2  f i (t)   i (t)]

i1
• Example: Signal with a function of two spatial coordinates in a plane
s(x, y)  3x  2xy  1 0 y 2

• Black & white image: light intensity wrt two spatial coordinates
• Video signals in TV: sequence of images (frames)
: two spatial coordinates and time
Examples
Types of Signals
Signals may be classified into four categories depending on the
characteristics of the time-variable and the values they can take:
Signals

Continuous-time Discrete-time

Continuous-value / Discrete-value Continuous-value Discrete-value

Analog Discrete Digital

Types of Signals
• Continuous time –
Continuous amplitude

• Continuous time –
Discrete amplitude

• Discrete time –
Continuous amplitude

• Discrete time –
Discrete amplitude
Continuous-time (analog) signals
Time: defined for every value of time t  R
Descriptions: functions of a continuous variable t → f (t)
Notes: they take on values in the continuous
interval f (t) (a,b) for a, b   .
Discrete-time signals:
Time: defined only at discrete values of time → t = nT
Descriptions: sequences of real/complex numbers → f (n)
Notes: they take on values in the continuous interval,
f (n) (a,b) for a, b  
Digital signals:
Descriptions: discrete-time and discrete-valued signals
(i.e. taking values from a finite set of possible values)
Note: sampling, quantizing and coding process i.e. process of analog-
to-digital conversion.

10
What is system?
• Entity that processes a set of signals
• Entity that manipulates one or more signals to accomplish a
function, thereby yielding new signals.
• Examples: software systems, electronic systems, computer systems,
or mechanical systems
What is system?
• Example (specific for the course): Filter system
• Performs operations on the signal to filter out the
noise/interference
• I.e. to reduce noise/interference corrupting the desired
signal
• OPERATIONS = SIGNAL PROCESSING
Analog versus Digital
Signal Processing

Analog Digital
V1(t)V1(t) V2 (t)
V2(t)

i(t)
i(t)

Digital processing on signal is performed by digital hardware

(logic circuit) and digital software program  digital system
Digital Signal Processing (DSP)
• Digital signal processing (DSP) includes the mathematics, the algorithms,
and the techniques used to manipulate these signals after they have
been converted into a digital form.
• Representation of signals by sequences of numbers or symbols AND the
processing of these sequences.
• Mathematical and algorithmic manipulation of discretized and quantized
or naturally digital signals in order to extract the most relevant and
pertinent information that is carried by the signal.
• Digital representation of signals and the use of digital processors to
analyse, modify, or extract information from signals.
Digital Signal Processing (DSP)
Refers to processing of signal using digital technology

Signal - A function or physical System - an Processing

quantity of a set of independent abstraction of By processing we mean operating
variables that contains some kind of anything that takes in a fashion on signal to extract
information which can be conveyed/ an input signal, some useful information
displayed/ manipulated. Examples: operates on it, and
Voltage, Image, Movie produces an output
signal
Digital Signal Processing (DSP)
Why process signals?
• Signals are corrupted by interference due to noise.
• Signals are distorted due to the transmission medium.
• Recover information present in the signal.
• Perform algorithms such as modulation, compression etc.
Components of a DSP System

Block diagram of a real-time digital signal processing system

Digitization
• As most of the signals encountered in science and engineering are
analog in nature, a process on converting the analog signal to discrete
signal is needed.
• This process is called digitization, which consists of Sampling and
Quantization process.
• When we sample, we calculate the value of the continuous signal at
discrete points
– How fast do we sample
– What is the value of each point
• Quantization determines the value of each samples value
Digitization
• Referring to figure below,
– converting continuous value to discrete value on horizontal-axis is called
sampling
– converting the continuous value to discrete value on vertical-axis is called
quantization.
Sampling Theorem
• Equation for the sampling process:

t represents continuous time (s), n represents the sample or discrete time

(integer) and is the sampling time (s).
• To ensure the digitized signal represents its original signal sufficiently
and correctly, the sampling process need to follow the Nyquist theorem:

represents the sampling frequency

represents the highest frequency component in the original signal.
DSP Operation: Examples
• Elementary time-domain operations
• Signal correlation
• Digital filtering
• Modulation and demodulation
• Discrete transformation from time domain to frequency
domain and vice versa
• Multiplexing and demultiplexing
DSP Operation: Filtering
 By far the most commonly used DSP operation
 Filtering deliberately change the frequency content of the signal,
typically, by removing certain frequencies from the signals.
 For denoising applications, the (frequency) filter removes those
frequencies in the signal that correspond to noise.
 In communications applications, filtering is used to focus to that part
of the spectrum that is of interest, that is, the part that carries the
information.
DSP Operation: Filtering
 Typically, we have the following types of filters
 Lowpass (LPF) – removes high frequencies, and retains (passes) low
frequencies
 Highpass (HPF) – removes low frequencies, and retains high frequencies
 Bandpass (BPF) – retains an interval of frequencies within a band,
removes others
 Bandstop (BSF) – removes an interval of frequencies within a band,
retains others
 Notch filter – removes a specific frequency
A Practical Example –
Low Pass Filtering (LPF)
- Sound:
• Clean:
• Noisy:
- Time Domain:
• Clean:
• Noisy:
- Frequency Domain:
• Clean:
• Noisy:

 Frequency Domain separates signal and noise

23
Applications of DSP
 Communication – Modulation and demodulation
 Signal security – Encryption and decryption
 Data compression – reduce space/computation required to store/process
data
 Signal denoising – Filtering for noise reduction
 Audio processing – Signal enhancement, equalization
 Image processing – Image denoising, enhancement, watermarking,
reconstruction
 Data analysis and feature extraction – Recognize structure on data
 Frequency/Spectral analysis – Alternate approach to time domain
analysis
Other Applications of DSP
 Image processing

Pattern recognition, Robotic vision, Image enhancement, Facsimile
 Telecommunications

Echo cancellation, Adaptive equalization, Video conferencing, Data
communication
 Biomedical

Patient monitoring, Scanners, ECG analysis (electrocardiography), EEG brain
mappers (electroencephalogram)
 Consumer appliances

Digital, cellular mobile phones, Digital TV, Digital cameras, CD/VCD/DVD
players/writers, PlayStation
Advantages of a DSP system
 Guaranteed accuracy: Determined by the no. of bits used.
 Perfect reproducibility: Digital recoding can be copied repeatedly without
degrading the signal quality.
 No drift in performance with temperature or age.
 Greater reliability, smaller size, and lower power consumption due to advances
in semiconductor technology.
 Greater flexibility: Can be programmed and reprogrammed without
modifications to the hardware.
 Superior performance: Can perform functions that are not possible with analog
methods, e.g. adaptive filtering, encryption, channel coding / decoding.
 Information is already in digital form (e.g. digital computer), DSP is the only
option.
Disadvantages of a DSP system
 Complexity: additional pre- and post-
devices (ADCs/DACs,processing
associated filters and
complex digital circuitry)
 Speed and cost: DSP designs and ADCs/DACs can be
expensive for broadband signals. Most DSP are still not
fast enough for wideband applications. Bandwidths in the
100 MHz range are still processed by analog means.
 Design time: Due to the shortage of knowledgeable DSP
engineers and the necessary resources (software
packages), DSP designs can be time consuming.
SUMMARY
DSP refers to processing of signals using digital technology.
It is a better alternative to analog signal processing.

Now you should know

 Digital signal processing (DSP) and the basic elements
in the system
 The reason why digital signal processing of analog
signals is preferred.
 The requirement of ADC and DAC as interfaces
between analog and digital domain

31
Thank You
DSP - Basic Mathematics
• Trigonometry
• Complex numbers
• Complex exponentials
• Factorisation
• Long division
• Irrational functions
• Sum/Product
• Functions: Change of variable
32
Trigonometr
y
sin
• Trigonometric circle: 1 

 1   
cos
 





 cos / 3  1/   
cos 0   sin  cos   
sin 0  sin  / 6  1/  
1 2  2
2 .
0
cos  1
cos / 4  2/2

sin   sin  / 4  2/ ..
 cos 0 
 0cos / 2  2 cos  cos
sin 0   cos / 6  3/2 sin  
1
0

0sin  / 2 

sin  / 3 3/2 .. 
sin
1  .
32
Complex numbers
93 2  1.41421... 1  i  j
42 3  1.73205... 4 (1)(4)  (1) 4
 2j
11
3.1415926535897932384626433832...  2  1.41421... j
00
 irrational numbers j 2  1

5  3 j : complex number
5  3 j2  4 j  10  20 j  6 j a  bjc  dj ac  db ad 
12 j 2 a :real
bc j part
 10 14 j 121 b : imaginary
 10 14 j 12 part
 22 14 j
Imaginary Im
plane:

Re

33
Complex exponentials
Imaginary Im Im
plane: 1

 1
Re Re

carthesian coordinates : polar

a  jb coordinates :
r
e j  e  j
Euler formula : e  cos  j
j
cos
sin 2
j
e  e
j
re j  rcos  j sin  r cos  jr sin
sin 2j
e j 2  ?e j
cos2  j sin 2  cos  j sin
2

cos   sin   j2 cos sin    cos   sin

2 2 2 2 2
  2 j cos 
34
sin 
Factorizatio
n
   f x a  bx  cx2
f x   x  x  x  x 
    D 2  b 2  4ac  discrimina

/, / : roots ( f x 0) nt
b D
f x      x  x i 2a  roots
x2 f x ax  x1 x 
x2 
D  0 : 2 different, real roots
D  0 : 2 equal, real roots
D  0 : 2 different, complex
roots

factorization : write f xin function of its

roots
f xof order n  f xhas n roots
 roots real or complex
 roots equal or different

order 1: evident
order 2 : easy
order 
3 :3 : numerical, not
analytical
difficult
35
Long Division
1254
5142741
41
1
04
8
2
2
22
2
05
177
164
13 x2 + 8x + 22
x3 + 5x2 - 2x + 5x - 3
x3 - 51427
3x2 / 41
=8x1254,
- 2x
2

8x -2
remainder
24x
13
22
x
+ 71
 -----------------
5 = (x + 8x + 22) + ------
2

x - 22 3 x 37
- 3x
Irrational Functions
3 25
rational numbers :1, 2, , ,...
4 36

irrational numbers : , e, 2, sin ,...
10
irrational functions : sin, cos, exp,
log,...

calculate : tables or power

3
series 
5  ...  n x 2n1
example : sin x  x x  x
3! n 2n
5! 10 1!
other irrational
functions : tan,arctan,
cosh,sinh,...
sin x
sincx
x
I0 x: zeroth order modified Bessel function of the first
n x: Jacobian elliptic function of order n
Ukind
38
5 Sum/Product
1 2  3  4  5   n : depends on 1, 5,
NOT n 7

n1

1.141. 1.9116
.  1
1  25  36  49   n 2
 n1 n n1
234 5 5
 1 1 1 1 1
0.5n  1      ...  ?

n 2 4 8 16
0 32


a n
 1 a
1 a
n
0
1

N 1 N2
1 a N a N1 1  a N 2
a n

1 a a
n

n0  n
N
1
1 a
N 1 N 1

(N 1)a  Na 
 na
N
n

n0 a
N 1 
1 (1 a) 2 1
n  a n

1 a
a
n0
N (N 1)
n
0
 1
N 1
21 
a
 1)(2N 1) 6N (N
n 2
  na n

(1 a) 2
a
n0 n
0
 1 39
Functions & variables
change variable  change each of its occurences

 

f x  x 2  x 1 sin x
 
4 
 evaluate : say x  2  f  
2 2 2 1 sin 2   7.1 


 2   7 4 
change variable : say x  5 y 
3  5 y  
2

 f 5 y  3 5 y  3  5 y  sin


 4 3 
31 
 25 y 2  35 y 13 sin y 
5

3
 4 
4 

N 1
1 a N
a n

1 a
n0 
1
5 1 65 1
4 51
1
n
1   1 5
6 16 6
5 5
7775
  1   
n
6  
6  6
5

1 1
6 1 616 1 6 .5
5 4
6480 1.1998457...
n0  6
 n0  6
 1 6 1 6 6
M N 1
1 a N
a n
a n

1 a
M  N 1  N  M 1
n
0M
? n
0

a
M 1
n 1 a
40
n
0
