Topic 1

Signals
Introduction
Signal: Any variable that carries or contains some kind of
information that can be conveyed, displayed or manipulated

System: A device that performs an operation on a signal to produce

desired output

Signal Processing: Operation on a signal to produce desired output

Digital Signal Processing: Concerned with the digital representation

of signals and the use of digital processors to analyze, modify and/or
extract information from it
Introduction
Application Areas of SP
Image processing: pattern recognition, robotic vision, image
enhancement, facsimile, satellite weather map, animation
Instrument and control: spectrum analysis, position and rate
control, noise reduction, data compression
Speech / audio: speech recognition, speech synthesis, text to speech,
digital audio, equalization
Military: secure communication, radar processing, sonar processing,
missile guidance
Telecommunications: echo cancellation, adaptive equalization,
modulation, spread spectrum, video conferencing, data
communication
Medical: patient monitoring, scanners, EEG (Electroencephalogram)
mapping, ECG (Electrocardiogram) analysis, X-ray storage /
enhancement
Where not?
Signals
Typical SP Operations
Elementary time-domain operations
 Scaling (amplification/attenuation): multiplication by a +ve
or –ve constant
 Time Shifting (delay / Advance): y(t) = x(t ± t0)
 Addition: y(t) = x1(t)+x2(t)-x3(t)+.......
 Integration
 Differentiation
Filtering: In addition to basic filters (LPF, HPF, BPF, BSF)
Notch filter: band stop filter designed to block a single frequency
Comb filter: designed to block frequencies that are integral
multiple of a certain low frequency
Multiband filter: more than one PB and more than one SB
Generation of signal (prediction)
Modulation and demodulation
Multiplexing and demultiplexing
Signal Types
Generating source: Single channel (scalar), Multichannel (vector)
Dimension (No. of dependent variable): 1-D, 2-D, 3-D.
Statistical parameters: Stationary, Non-stationary
Real, imaginary or complex
Certainty of description:
Deterministic signal: can be uniquely determined by a well-
defined process such as mathematical expression or rule, or
table look-up, etc. Deterministic signal can be linear or
nonlinear.
Random signal: is generated in a random fashion and can not
be predicted ahead of time.
Signal Types
Periodicity: Periodic, Aperiodic
Summation: Energy, Power
Continuity of independent variable (t) and signal:
• Continuous time (CT) or Analog signal: a continuous-time
signal with continuous magnitude
• Discrete time (DT) or Sampled signal: discrete-time signal
with continuous-valued magnitude
• Discrete or Quantized signal: discrete-time signal with
discrete-valued magnitude
• Digital signal: a discrete-time signal with discrete-valued
magnitudes represented by a finite number of digits.
Signal Examples

Electrical signals
– Voltages and currents in a circuit
Acoustic signals
– Acoustic pressure (sound) over time
Mechanical signals
– Velocity of a car over time
Video signals
– Intensity level of a pixel (camera, video) over
time
Signals in an Electrical
Circuit
R vs (t )  vc (t )
i (t ) 
R
dv (t )
+ i i (t ) C c
vs C vc dt
-
dvc (t ) 1 1
 vc (t )  vs (t )
dt RC RC
The signals vc and vs are patterns of variation over time

Step (signal) vs at t=1

vs, vc

RC = 1 (yellow)
First order (exponential) response
for vc (red)

Note, we could also have considered the voltage across the resistor or the
current as signals
Signal Classification

Type of Independent Variable

Cervical MRI
Independent Variable Dimensionality
Continuous Time (CT) and Discrete-Time (DT) Signals

14
Continuous Time (CT) Signals
A signal is continuous if at any time t1, x(t1  )  x(t1 )  x(t1 ) .

Piecewise continuous signals

(a) Rectangular pulse
 
1 t 
x(t )  A rect (t /  ) rect (t /  )  2

0 t 
 2

Continuous everywhere except at t 
2
 Continuous Time (CT) Signals

(a) Pulse train: Continuous at all t except at

t kTs k is any integer

At any point of discontinuity, the signal is assigned the

value of
1
x(t1 )  x(t1 )  x(t1  )
2
DT signals only occur at discrete values of time, say, at kTs.
Periodic and Aperiodic Signals
Periodic Signals
A signal is said to be periodic if x(t )  x(t  nT ) n 0,1, 2,.. ...
1 2
T = fundamental period = 
f 0 0
Any frequency  k k 0 is called harmonics

Figure: Harmonically related sinusoids

If two signals are periodic; x(t) with period T1 and y(t) with T2 and
let z(t) = ax(t) + by(t). To find the condition of periodicity of z(t), we
have
x(t )  xt  kT1  y (t )  y t  lT2 
If z(t) is periodic with the period T, then
z (t )  z (t  T ) ax(t  T )  by (t  T ) axt  kT1   by t  lT2 
This condition is valid if we have only
T l
T kT1 lT2  1  rational number
T2 k
Energy and Power Signals

Energy and power of a signal are defined as

Lim L 2 Lim  1 L 2 
L    L    2 L 
E x (t ) dt P  x (t ) dt 
L L 
If the limit exists and 0  E   then the signal is energy signal,
and if 0  P   , then power signal. The energy signal has 0
power while power signal has  energy.

If a signal is periodic with period T, then 2L=mT and

Lim  1 T 2  1T 2

m    mT  
P  m x (t ) dt   x (t ) dt
0  T 0
Energy and Power Signals
Example 1: Let x(t )  Ae  t t 0

A2
E  A e2  2t
dt   Energy signal
0
2
Lim  1 L 2  2t  Lim A 2
L    2 L 
P  A e dt   0
L  L   4L

A t 0
Example 2: x(t )   t
 Ae t 0
Lim  0 2 L
 Lim 2  1 
E  A dt  A e dt  
L   L
2  2t

A  L  1  e  2 L  
0  L   2 
 not energy signal
Lim 1 2  1  A
2

 L  1  e    Power signal
 2L
P A
L   2L  2  2
Real and Complex Signals
Right- and Left-Sided Signals
Bounded and Unbounded Signals
Even and Odd Signals
Common Elementary Signals
1 t 0
Unit step function u(t) x(t ) 
0 t 0
(a) Rectangular pulse
 
1 t 
x(t )  A rect (t /  ) rect (t /  )  2

0 t 
 2
1 t 0

(b) Signum function sgn(t) x(t ) 0 t 0
 1 t 0

t t 0
Ramp function r(t) x(t ) 
0 t 0
Unit impulse function (t )
(0)   (t ) 0 for t 0


(t ) dt 1

(t ) ( t )

Other common CT signals are:

sin t
i. Sampling function Sa (t ) 
t
sin  t
ii. Sinc function sin c(t ) 
t
Unit Impulse Function
Narrow Pulse Approximation
Intuiting Impulse Definition
Uses of the Unit Impulse
Unit Step Function
Successive Integrations of the Unit Impulse Function
Elementary SP Operations
Time Shifting x(t )  x(t   )

Figure: Shifting operation

t  1  1 t 0
1 0  t 2

Example: Let x(t ) 
  t 3 2  t 3
0 otherwise
Elementary SP Operations
Folding / Reflection / Image x(t )  x( t )

Example: Folding and time shifting

t  1  1 t 0

Let x(t )  1 0  t 2
0 otherwise

Elementary SP Operations

Time Scaling x(t )  x(at )

 Elementary SP Operations
Relation between elementary signals
Any other elementary signal can be obtained from unit step
function.
 t 
Rectangular pulse: A rect    Au (t  a)  u (t  a)
 2a 
t

Ramp function:
r (t )  u ( )d

d
Impulse function: (t )  u (t )
dt
Properties of Impulse function
t2
 x(t 0 ) t1  t  t 2
Shifting 
t1
x (t ) (t  t 0 ) dt  
0 Otherwise

x(t )  x( )(t   )d

 any signal can be expressed as a continuous summation
(integration) of weighted impulses
Building Block Signals can be used to create a rich variety of
Signals
Digital Signal
A discrete‑time signal having a set of discrete values
Analog signal

Discrete-time (DT)
signal

Discrete signal
Digital Signal
Advantages
Guaranteed accuracy, Perfect reproducibility (High Fidelity)
No drift of performance with temperature or age
High storage capability
Superior performance

Limitations
Speed and cost
Design time
Finite word length problems
But these limitations are being continually diminished with the advent
of new technology
A/D Conversion
Sampling, Quantization, Encoding
Sampling
Conversion of a continuous-time signal into a discrete time signal
obtained by taking “samples” of the continuous-time signal at
discrete-time instants
T  sampling interval, sec
Fs (=1/T)  sampling frequency (sample/sec, Hz)
Sampling

Continuous-time signal Discrete-time signal

Ω = 2πF ω = 2πf
Ω  radian/sec ω  radian/sample
F  cycle/sec (Hz) f  cycles/sample (Hz)
F can have any value f is a rational number
Rate of oscillation increases with Rate of oscillation increases for a
Ω or F change of ω of 2π
-∞ < Ω < ∞ -π ≤ ω ≤ π
-∞ < F < ∞ -½ ≤ f ≤ ½
f = F/Fs
Fs  sample/sec
Sampling Example: Aliasing

x1 (t ) cos 2 (10)t sampled at a rate Fs = 40 Hz

x 2 (t ) cos 2 (50)t

x1 (n) cos 2 ( 10
40 ) n cos 
2 n

x2 (n) cos 2 ( 50
40 ) n cos 5
2 n cos 
2 n

Nyquist Sampling Theorem: Fs 2Fmax

Nyquist rate:
Fs 2Fmax
Quantization
Discrete leveling of DT signal (Approximation: rounding and
truncation, ceiling)
Quantization
Quantization with one significant digit using Rounding
Discrete-time xq(n) Error
n signal (Rounding) eq(n) = xq(n)-x(n)
x(n)
0 1 1.0 0.0
1 0.9 0.9 0.0
2 0.81 0.8 -0.01
3 0.729 0.7 -0.029
4 0.6561 0.7 0.0439
5 0.59049 0.6 0.00951
6 0.531441 0.5 -0.031441
7 0.4782969 0.5 0.0217031
8 0.43046721 0.4 -0.03046721
9 0.387420489 0.4 0.012579511
Quantization: Sinusoid
Quantization Noise
xmax  xmin b
 
2 eq (t )  
2
 L L 2
 

 e (t ) dt  e (t ) dt
2 2
Pq  1
2 q
1
q eq (t )  t   t 
 0 2

2A
Pq   ( ) t dt 
1 
2
2 2 2
12
 b
0
2
A2 / 3
Pq  Tp

( A cos  t ) dt  A2
2
22 b Ps  1
Tp 0 2
ps 2b 0
SQNR   . 2 Pq
3
2

SQNR(dB) 10 log10 SQNR 1.76  6.02b

SQNR increases approximately at the rate of 6 dB/bit
Encoding
Representation of discrete (quantized) signals by symbols
Mostly used is binary [0 1]
Lowest value by 000000…
Highest value may be 1 1 1 1 1 1 …

Representation of Discrete Time (DT) Signal:

x(n) = {…………………}

Some typical elementary DT Signals are:

Unit step function: x(n) = u(n)
Impulse function: x(n) = δ(n)
Ramp function: x(n) = r(n)=n
Power function: x(n) = an
Typical DT SP Operations
Elementary time-domain operations
Scaling (amplification/attenuation): x(n)  ax(n)
Time shifting (delay / advance): x(n) x(n ± k)
Folding / image / reflection: x(n)  x(-n)
Caution! x(-n-k) ≠ x[-(n+k)]
Time-scaling: x(n)  x(an)
Addition: y(n) = x1(n)+x2(n)-x3(n)+.......
Subtraction, Multiplication, Division, Integration, Differentiation
Thank you