Nguyễn Công Phương

SIGNAL PROCESSING

Discrete – Time Signals and Systems

 Contents
I. Introduction
II. Discrete – Time Signals and Systems
III. The z – Transform
IV. Fourier Representation of Signals
V. Transform Analysis of LTI Systems
VI. Sampling of Continuous – Time Signals
VII.The Discrete Fourier Transform
VIII.Structures for Discrete – Time Systems
IX. Design of FIR Filters
X. Design of IIR Filters
XI. Random Signal Processing
s i tes.google.com/site/ncpdhbkhn 2
 Discrete – Time Signals and
Systems
1. Discrete – Time Signals
2. Discrete – Time Systems
3. Convolution Description of Linear Time – Invariant
Systems
4. Properties of Linear Time – Invariant Systems
5. Analytic Evaluation of Convolution
6. Numerical Computation of Convolution
7. Real – Time Implementation of FIR Filters
8. FIR Spatial Filters
9. Systems Described by Linear Constant – Coefficient
Difference Equations
10. Continuous – Time LTI Systems

s i tes.google.com/site/ncpdhbkhn 3
 Discrete – Time Signals (1)
• x[n]: a sequence of numbers x[n]
defined for every value of the
integer variable n T

• It is not defined for noninteger

value of n 1 2 5 6 n
N2 0
• {x[n ]}N1 :the samples in the 3 4 7 8
x (t )
range from N1 to N2
• T (sampling period): the x[n] = x(t ) t = nT = x (nT )
interval between two
successive samples, measured
in seconds (s)
• fs (sampling frequency, 1/T):
the number of samples per unit
of time, measured in hertz (Hz)
s i tes.google.com/site/ncpdhbkhn 4
 Discrete – Time Signals (2)

0, n<0

x[n ] =  1 n … –2 –1 0 1 2 …
 3n , n≥0 x[n] … 0 0 1 1/3 1/9 …

Function Table
Sequence Graph
1

0.9

0.8

0.7

0.6

1 1 1
x[n ] = {... 0 1
0.5

...} 0.4

↑ 3 9 27
0.3

0.2

0.1

0 0 0.5 1 1.5 2 2.5 3 3.5 4

s i tes.google.com/site/ncpdhbkhn 5
Discrete – Time Signals (3)


2
Energy: Ex = x[ n ]
n =−∞

 1 L
2
Power: Px = lim 
L →∞ 2 L + 1


n =− L
x[n ] 


s i tes.google.com/site/ncpdhbkhn 6
 Discrete – Time Signals (4)
δ [ n] δ [n − 3]
1

0.8

0.6

0.4

0.2
n
0
-2 -1 0 1 2 3 4

1, n=0
Unit sample sequence (impulse): δ [n] = 
0, n≠0
1, n=k
δ [n − k ] = 
0, n≠k
s i tes.google.com/site/ncpdhbkhn 7
 Discrete – Time Signals (5)
1

0.8

0.6

0.4

0.2
n
0
-2 -1 0 1 2 3 4

1, n≥0
Unit step sequence: u[n ] = 
0, n<0

s i tes.google.com/site/ncpdhbkhn 8
 Discrete – Time Signals (6)

n
0

Sinusoidal sequence: x[n ] = A cos(ω0n + φ ), −∞<n<∞

s i tes.google.com/site/ncpdhbkhn 9
Discrete – Time Signals (7)

Exponential sequence: x[n ] = Aa n , 0 < a <1

s i tes.google.com/site/ncpdhbkhn 10
 Discrete – Time Signals (8)

0
n
-1

-2

-3

-4
0 1 2 3 4 5 6 7 8 9 10

Exponential sequence: x[n ] = Aa n , −1< a < 0

s i tes.google.com/site/ncpdhbkhn 11
 Discrete – Time Signals (9)
1

0.8

0.6

0.4

0.2

0 n
-0.2

-0.4

-0.6

-0.8

-1
0

Periodic sequence: x[n ] = x[n + N ]

s i tes.google.com/site/ncpdhbkhn 12
 Discrete – Time Signals and
Systems
1. Discrete – Time Signals
2. Discrete – Time Systems
3. Convolution Description of Linear Time – Invariant
Systems
4. Properties of Linear Time – Invariant Systems
5. Analytic Evaluation of Convolution
6. Numerical Computation of Convolution
7. Real – Time Implementation of FIR Filters
8. FIR Spatial Filters
9. Systems Described by Linear Constant – Coefficient
Difference Equations
10. Continuous – Time LTI Systems

s i tes.google.com/site/ncpdhbkhn 13
 Discrete – Time Systems (1)

x[n]
H y[n]
Discrete – time
system
H
x[n] ֏ y[n ]
y[n] = H {x[n]}
• Causality
• Stability
• Linearity
• Time – invariance
s i tes.google.com/site/ncpdhbkhn 14
 Discrete – Time Systems (2)
• Definition: A system is called causal if the
present value of the output does not depend on
future values of the input
• That is, y[n0] is determined by the values of
x[n] for n ≤ n0, only
• Ex. 1: y[n] = x[n] + 2x[n – 1] + x[n – 2]
• Ex. 2: y[n] = x[n – 1] + x[n] + x[n + 1]

s i tes.google.com/site/ncpdhbkhn 15
 Discrete – Time Systems (3)
• Definition: A system is called stable, in the
Bounded-Input Bounded-Output (BIBO)
sense, if every bounded input signal results in
a bounded output signal
• That is, |x[n]| ≤ Mx < ∞  |y[n]| ≤ My < ∞
• A signal x[n] is bounded if there exists a
positive finite constant Mx such that |x[n]| ≤ Mx
for all n
• Ex. 3: y[n] = x[n] + 2x[n – 1] + x[n – 2]
• Ex. 4: y[n] = 2nx[n]
s i tes.google.com/site/ncpdhbkhn 16
 Discrete – Time Systems (4)
• Definition: A system is called linear if and
only if for every real or complex constant a1,
a2 and every input signal x1[n] and x2[n]:

H{a1x1[n] + a2x2[n]} = a1H{x1[n]} + a2H{x2[n]}

• A.k.a. the principle of superposition

• Ex. 5: y[n] = 2x[n]
• Ex. 6: y[n] = x2[n]

s i tes.google.com/site/ncpdhbkhn 17
 Discrete – Time Systems (5)
• Definition: A system is called time – invariant or
fixed if and only if:

y[n] = H{x[n]}  y[n – n0] = H{x[n – n0]}

for every input x[n] and every time shift n0

• That is, a time shift in the input results in a
corresponding time shift in the output
• Ex. 7: y[n] = 3x[n]
• Ex. 8: y[n] = x[n]/n
s i tes.google.com/site/ncpdhbkhn 18
 Discrete – Time Systems (6)
x 2 [n ]

y[n] = x1[n] + x2 [n]

x1[n ] +
Adder
Summing node x2 [n]

y[n] = x1[n ] + x2 [n]

x1[n ]

s i tes.google.com/site/ncpdhbkhn 19
 Discrete – Time Systems (7)

a y[n] = ax[n ]
x[ n]

Multiplier
Gain branch

a y[n] = ax[n ]
x[ n]

s i tes.google.com/site/ncpdhbkhn 20
 Discrete – Time Systems (8)

−1
y[n] = x[n − 1]
x[ n] z

Unit delay
Unit delay branch

z −1 y[n] = x[n − 1]
x[ n]

s i tes.google.com/site/ncpdhbkhn 21
 Discrete – Time Systems (9)
w[n] w[n]

w[n]
Splitter
Pick-off node

w[n] w[n]

w[n]
s i tes.google.com/site/ncpdhbkhn 22
 Discrete – Time Systems (10)
Ex. 8 w[n ] a y[n ]
x[n ]
Find y[n]?

z −1

c b

w[n] = x[n] + cw[n − 1] bx[n ] + cy[n ]

→ w[n ] =
y[n] = aw[n] + bw[n − 1] b + ac
bx[n − 1] + cy[n − 1]
→ w[n − 1] =
b + ac
= a( x[n] + cw[n − 1]) + bw[n − 1] = ax[n] + (ac + b)w[n − 1]

bx[n − 1] + cy[n − 1]
→ y[n ] = ax[n] + (ac + b) = ax[n ] + bx[n − 1] + cy[n − 1]
b + ac
s i tes.google.com/site/ncpdhbkhn 23
 Discrete – Time Systems (11)
• Discrete – time system properties:
– Causality: x[n ] = 0 for n ≤ n0 → y[n ] = 0 for n ≤ n0
– Stability: x[n] ≤ M x < ∞ → y[n ] ≤ M < ∞
– Linearity:  k ck xk [n] →  k ck yk [n]
– Time – invariance: x[n − n0 ] → y[n − n0 ]
• Practical realizability:
– A finite amount of memory for the storage of
signal samples and
– A finite number of arithmetic operations for the
computation of each output sample

s i tes.google.com/site/ncpdhbkhn 24
 Discrete – Time Signals and
Systems
1. Discrete – Time Signals
2. Discrete – Time Systems
3. Convolution Description of Linear Time –
Invariant (LTI) Systems
4. Properties of Linear Time – Invariant Systems
5. Analytic Evaluation of Convolution
6. Numerical Computation of Convolution
7. Real – Time Implementation of FIR Filters
8. FIR Spatial Filters
9. Systems Described by Linear Constant – Coefficient
Difference Equations
10. Continuous – Time LTI Systems

s i tes.google.com/site/ncpdhbkhn 25
 Convolution Description of LTI
Systems (1)
• A great need for the evaluation of the performance of
an LTI system
• The response of a system can be used to evaluate its
performance
• This can be obtained from impulse responses h[n] by
using the convolution

1 1
δ [n ] h[n]
0 LTI system 0
Unit impulse Impulse response

s i tes.google.com/site/ncpdhbkhn 26
Convolution Description of LTI
Systems (2)
Convolution
x[n] H y[n] = ?
LTI system

H
LTI system

H
→ y[ n]
LTI system

H
LTI system
s i tes.google.com/site/ncpdhbkhn 27
 Convolution Description of LTI
Systems (3)
 x[k ], n = k
xk [n] = 
0, n≠k
0 1 2

x[0]δ [n]
0 1 2

x[1]δ [n − 1] → x[n] = x[0]δ [ n] + x[1]δ [n − 1] + x[2]δ [n − 2]

0 1 2
∞
x[n ] =  x[k ]δ [n − k ]
k =−∞
x[2 ]δ [n − 2 ]
0 1 2
s i tes.google.com/site/ncpdhbkhn 28
 Convolution Description of LTI
1
Systems (4) 1
δ [n ] h[n]
0 LTI system 0

1 1
δ [n − k ] h[n − k ]
k LTI system k

x[k] x[k]
x[k ]δ [ n − k ] x[k ]h[n − k ]
k LTI system k

∞
y[n] =  x[k ]h[n − k ] = x[n ]* h[n ],
k =−∞
− ∞ < n < ∞ (Convolution)
s i tes.google.com/site/ncpdhbkhn 29
 Convolution Description of LTI
Systems (5)
∞
y[n] =  x[k ]h[n − k ] = x[n]* h[n], −∞<n<∞
k =−∞

y[−1] = ... + x[ k ]h[n − k ] n= −1,k =−1 + x[k ]h[n − k ] n =−1,k =0 +

+ x[ k ]h[n − k ] n= −1,k =1 + x[k ]h[ n − k ] n=−1,k =2 ...
= ... + x[−1]h[0] + x[0]h[−1] + x[1]h[−2] + x[2]h[ −3] + ...
y[0] = ... + x[k ]h[n − k ] n =0,k =−1 + x[ k ]h[n − k ] n=0,k =0 +
+ x[ k ]h[n − k ] n=0,k =1 + x[ k ]h[n − k ] n=0,k =2 ...
= ... + x[−1]h[1] + x[0]h[0] + x[1]h[ −1] + x[2]h[−2] + ...
y[1] = ... + x[ k ]h[n − k ] n=1,k =−1 + x[k ]h[ n − k ] n=1,k =0 +
+ x[k ]h[n − k ] n=1,k =1 + x[k ]h[n − k ] n =1,k =2 ...
= ... + x[−1]h[2] + x[0]h[1] + x[1]h[0] + x[2]h[ −1] + ...
s i tes.google.com/site/ncpdhbkhn 30
 Convolution Description of LTI
Ex. 9 Systems (6)
Given x[n] = {1 2 3 4 5}, h[ n] = {−1 2 1}. Find x[n]*h[n]?
↑ ↑

y[0] = 0 +2 −2 = 0
0 0 1 2 3 4 5 0 0
× × ×
0 1 2 –1 0

y[1] = 1 +4 −3 = 2
0 0 1 2 3 4 5 0 0
× × ×
0 1 2 –1 0

y[ 2 ] = 2 +6 −4 = 4
0 0 1 2 3 4 5 0 0
× × ×
0 1 2 –1 0
s i tes.google.com/site/ncpdhbkhn 31
 Convolution Description of LTI
Ex. 9 Systems (7)
Given x[n] = {1 2 3 4 5}, h[ n] = {−1 2 1}. Find x[n]*h[n]?
↑ ↑

y[0] = −2 +2 0 = 0
0 0 5 4 3 2 1 0 0
× × ×
0 –1 2 1 0

y[1] = −3 +4 1 = 2
0 0 5 4 3 2 1 0 0
× × ×
0 –1 2 1 0

y[ 2 ] = −4 +6 2 = 4
0 0 5 4 3 2 1 0 0
× × ×
0 –1 2 1 0
s i tes.google.com/site/ncpdhbkhn 32
Convolution Description of LTI
Systems (8)

∞ ∞
y[n] =  x[k ]h[n − k ] =  h[k ]x[n − k ]
k =−∞ k =−∞

y[n ] = x[n ] * h[n ] = h[n ] * x[n ]

s i tes.google.com/site/ncpdhbkhn 33
 Discrete – Time Signals and
Systems
1. Discrete – Time Signals
2. Discrete – Time Systems
3. Convolution Description of Linear Time – Invariant Systems
4. Properties of Linear Time – Invariant Systems
a) Properties of Convolution
b) Causality and Stability
c) Convolution of Periodic Sequences
d) Response to Simple Test Sequences
5. Analytic Evaluation of Convolution
6. Numerical Computation of Convolution
7. Real – Time Implementation of FIR Filters
8. FIR Spatial Filters
9. Systems Described by Linear Constant – Coefficient Difference
Equations
10. Continuous – Time LTI Systems

s i tes.google.com/site/ncpdhbkhn 34
 Properties of Convolution (1)
x[ n] x[ n] x[ n] x[n − n0 ]
δ[n] δ[n – n0]

x[ n ] = δ [ n ] * x[ n ] x[n − n0 ] = δ [n − n0 ]* x[n ]

x[ n] y[ n ] h[ n ] y[ n ]
h[n] x[n]
y[n] = x[n]* h[n] = h[n]* x[n]

x[ n ] y[ n ] x[ n] y[ n ]
h1[n] h2[n] h[n] = h1[n]*h2[n]

( x[n] * h1[n ]) * h2[n ] = x[n] * ( h1[n] * h2[n ])

s i tes.google.com/site/ncpdhbkhn 35
 Properties of Convolution (2)
x[n] y[n] x[n] y[n]
h1[n] h2[n] h2[n] h1[n]

( x[n]* h1[n ]) * h2[n ] = ( x[n]* h2 [n]) * h1[n]

h1[n]
x[n] y[n] x[ n] y[ n ]
+ h[n] = h1[n]+h2[n]

h2[n]

x[n] * (h1[n ] + h2 [n]) = x[n] * h1[n] + x[n ]* h2 [n]

s i tes.google.com/site/ncpdhbkhn 36
 Causality and Stability
• A linear time – invariant system with impulse
response h[n] is causal if:
h[ n] = 0 for n<0

• A linear time – invariant system with impulse

response h[n] is stable, in the bounded – input
bounded – output sense, if and only if the impulse
response is absolutely summable, that is, if:
∞

 h[n] < ∞
n =−∞
s i tes.google.com/site/ncpdhbkhn 37
 Discrete – Time Signals and
Systems
1. Discrete – Time Signals
2. Discrete – Time Systems
3. Convolution Description of Linear Time – Invariant Systems
4. Properties of Linear Time – Invariant Systems
a) Properties of Convolution
b) Causality and Stability
c) Convolution of Periodic Sequences
d) Response to Simple Test Sequences
5. Analytic Evaluation of Convolution
6. Numerical Computation of Convolution
7. Real – Time Implementation of FIR Filters
8. FIR Spatial Filters
9. Systems Described by Linear Constant – Coefficient Difference
Equations
10. Continuous – Time LTI Systems

s i tes.google.com/site/ncpdhbkhn 38
Convolution of Periodic
Sequences
∞
y[n] =  x[k ]h[n − k ]
k =−∞

∞
→ y[n + N ] =  x[ k ]h[n + N − k ]
k =−∞

x[n] = x[n + N ]

→ x[n − k ] = x[n + N − k ]
∞
→ y[n + N ] =  x[k ]h[n − k ]
k =−∞

→ y [ n + N ] = y [ n]
s i tes.google.com/site/ncpdhbkhn 39
 Response to Simple Test
Sequences (1)
∞ n
The step response: s[n ] =  h[k ]u[n − k ] =  h[ k ]
k =−∞ k = −∞
0
s[ 0] = u[0]h[0] = 1 × h[0] = h[ 0] =  h[k ]
h[ n]
k=0

s[1] = u[0]h[ 0] + u[1]h[1] = h[0] + h[1]

1
=  h[k ] 0
k=0
u[n]
s[ 2] = u[ 0]h[0] + u[1]h[1] + u[ 2]h[ 2] 1
2
= h[0] + h[1] + h[2] =  h[k ]
k=0
0
s[ 3] = u[0]h[0] + u[1]h[1] + u[ 2]h[2] + u[3]h[ 3]
3
= h[0] + h[1] + h[2] + h[3] =  h[k ]
k=0

s i tes.google.com/site/ncpdhbkhn 40
 Response to Simple Test
Sequences (2)
∞ n
The step response: s[n ] =  h[k ]u[n − k ] =  h[ k ]
k =−∞ k = −∞

s[ 0] = u[0]h[0] = 1 × h[0] = h[ 0]
→ s[1] − s[0] = h[1]
s[1] = u[0]h[ 0] + u[1]h[1] = h[0] + h[1]
s[ 2] = u[ 0]h[0] + u[1]h[1] + u[ 2]h[ 2] → s[2] − s[1] = h[2]

= h[0] + h[1] + h[2]

s[ 3] = u[0]h[0] + u[1]h[1] + u[ 2]h[2] + u[3]h[ 3] → s[3] − s[2] = h[3]
= h[0] + h[1] + h[2] + h[3]

h[n] = s[n] − s[ n − 1]

s i tes.google.com/site/ncpdhbkhn 41
 Response to Simple Test
Sequences (3)

x[n ] = δ [n ] ֏ y[n] = h[n]

n
x[n ] = u[n ] ֏ y[n] = s[n] =  h[k ]
k =−∞

x[ n ] = a n ֏ y [n ] = H ( a )a n

x[n ] = e jωn ֏ y[n] = H (e jωn )e jω n

∞
H (a ) = 
k =−∞
h[k ]a − k

s i tes.google.com/site/ncpdhbkhn 42
 Discrete – Time Signals and
Systems
1. Discrete – Time Signals
2. Discrete – Time Systems
3. Convolution Description of Linear Time – Invariant
Systems
4. Properties of Linear Time – Invariant Systems
5. Analytic Evaluation of Convolution
6. Numerical Computation of Convolution
7. Real – Time Implementation of FIR Filters
8. FIR Spatial Filters
9. Systems Described by Linear Constant – Coefficient
Difference Equations
10. Continuous – Time LTI Systems

s i tes.google.com/site/ncpdhbkhn 43
 Analytic Evaluation of
Convolution
h[ n], n ∈ [ M1, M 2 ]; x[n ], n ∈ [ N1, N 2 ]
→ y[ n], n ∈ [ M 1 + N 1, M 2 + N 2 ] h[ n − k ] x[ k ]
Partial overlap (left): n − M2 n − M1 N1 N2 k
n − M 1 ≥ N 1 n ≥ N 1 + M 1
 →
n − M 2 ≤ N 1 n ≤ N 1 + M 2
h[ n − k ] x[ k ]
n −M 1
y[ n ] =  x[k ]h[n − k ] ,
k =N 1
for N 1 + M1 ≤ n ≤ N1 + M 2
N1
Full overlap: n − M2 n − M1 N2 k
n − M 2 > N1 n > N1 + M 2
 →
n − M < N n < N 2 + M1
h[ n − k ]
1 2
x[ k ]
n −M 1
y[ n ] = 
k =n − M 2
x[k ]h[n − k ] , for N1 + M 2 < n < N 2 + M1
N2
Partial overlap (right): N1 n − M 2 n − M1 k
n − M 1 ≥ N 2 n ≥ M 1 + N2
 → 
n − M 2 ≤ N 2  n ≤ M 2 + N2
x[k ] h[ n − k ]
N2
y[n ] = 
k =n −M 2
x[k ]h[n − k ] , for N2 + M 1 ≤ n ≤ N 2 + M 2
N2
s i tes.google.com/site/ncpdhbkhn N1 n − M2 n− M
44
1 k
 Discrete – Time Signals and
Systems
1. Discrete – Time Signals
2. Discrete – Time Systems
3. Convolution Description of Linear Time – Invariant
Systems
4. Properties of Linear Time – Invariant Systems
5. Analytic Evaluation of Convolution
6. Numerical Computation of Convolution
7. Real – Time Implementation of FIR Filters
8. FIR Spatial Filters
9. Systems Described by Linear Constant – Coefficient
Difference Equations
10. Continuous – Time LTI Systems

s i tes.google.com/site/ncpdhbkhn 45
 Numerical Computation of
Convolution (1)
h[ n]
No overlap y[ −1] = h[0]0 + h[1]0 + h[2]0
y[0] = h[0] x[0] + h[1]0 + h[ 2]0
Partial overlap
y[1] = h[0]x[1] + h[1]x[0] + h[2]0 0
y[ 2] = h[0]x[2] + h[1]x[1] + h[2] x[0] x[n]
y[3] = h[0]x[3] + h[1]x[2] + h[2] x[1]
Full overlap
y[ 4] = h[0]x[4] + h[1] x[3] + h[2] x[ 2]
y[5] = h[0]x[5] + h[1]x[4] + h[2]x[3] 0
y[6] = h[0]0 + h[1]x[5] + h[2]x[4]
Partial overlap
y[7] = h[0]0 + h[1]0 + h[ 2] x[5]
No overlap y[8] = h[0]0 + h[1]0 + h[ 2]0

s i tes.google.com/site/ncpdhbkhn 46
 Numerical Computation of
Convolution (2)
 y [0 ]  x[0] 0 0 
 y[1]   x[1] x[0] 0 
   
No overlap y[ −1] = h[0]0 + h[1]0 + h[2]0  y[ 2]  x[2] x[1] x[0]
    h[0]
 y [ 3] =  x [ 3] x [ 2 ] x [1 ]   h[1] 
y[0] = h[0] x[0] + h[1]0 + h[ 2]0
Partial overlap  y[ 4]  x[4] x[3] x[2]  
y[1] = h[0]x[1] + h[1]x[0] + h[2]0     h[2]
 y [ 5]   x [5 ] x [ 4 ] x [ 3] 
y[ 2] = h[0]x[2] + h[1]x[1] + h[2] x[0]  y [6 ]  0 x[5] x[4]
   
y[3] = h[0]x[3] + h[1]x[2] + h[2] x[1]  y [7 ]  0 0 x[5]
Full overlap
y[ 4] = h[0]x[4] + h[1] x[3] + h[2] x[ 2]  x[0]  0   0 
 x[1]   x[0]  0 
y[5] = h[0]x[5] + h[1]x[4] + h[2]x[3]      
 x[ 2]  x[1]   x[0]
y[6] = h[0]0 + h[1]x[5] + h[2]x[4]      
Partial overlap x [ 3] x[ 2 ] x [1]
y[7] = h[0]0 + h[1]0 + h[ 2] x[5] = h[0]   + h[1]   + h [2 ]  
 x[ 4]  x[3]  x[2]
No overlap y[8] = h[0]0 + h[1]0 + h[ 2]0      
 x [ 5]   x[ 4 ]   x[ 3] 
 0   x[5]  x[4]
     
 0   0   x[ 5] 
s i tes.google.com/site/ncpdhbkhn 47
 Discrete – Time Signals and
Systems
1. Discrete – Time Signals
2. Discrete – Time Systems
3. Convolution Description of Linear Time – Invariant
Systems
4. Properties of Linear Time – Invariant Systems
5. Analytic Evaluation of Convolution
6. Numerical Computation of Convolution
7. Real – Time Implementation of FIR Filters
8. FIR Spatial Filters
9. Systems Described by Linear Constant – Coefficient
Difference Equations
10. Continuous – Time LTI Systems

s i tes.google.com/site/ncpdhbkhn 48
 Real – Time Implementation of
FIR Filter

Signal memory
In
0
x[0]
x[1]
x[2]
x[3] 0
x[0]
x[1]
x[2] 0
x[0]
x[1] 0
x[0]

Out
+ + + + h[0]0x[+13
0]]h++[1hh][0[11]+]x0xh[[+
2 12]]h]++[02hh+][[02h2]+][0x3xh[]+[0
0[2 1]3h]+[03h][03]0x[0]

h[0] h[1] h[2] h[3]

Coefficient memory

s i tes.google.com/site/ncpdhbkhn 49
 Discrete – Time Signals and
Systems
1. Discrete – Time Signals
2. Discrete – Time Systems
3. Convolution Description of Linear Time – Invariant
Systems
4. Properties of Linear Time – Invariant Systems
5. Analytic Evaluation of Convolution
6. Numerical Computation of Convolution
7. Real – Time Implementation of FIR Filters
8. FIR Spatial Filters
9. Systems Described by Linear Constant – Coefficient
Difference Equations
10. Continuous – Time LTI Systems

s i tes.google.com/site/ncpdhbkhn 50
 FIR Spatial Filters
1
y[ m, n] = ( x[m − 1, n − 1] + x[ m − 1, n] + x[ m − 1, n + 1]
9
+ x[ m, n − 1] + x[m, n ] + x[ m, n + 1]
+ x[ m + 1, n − 1] + x[m + 1, n] + x[ m + 1, n + 1])
1 1
1
=   9 x [m − k , n − l ]
k =−1 l =− 1

1
 , − 1 ≤ m, n ≤ 1 H
h[m, n ] =  9
0, otherwise
1 1
→ y[ m, n] =   h[ k, l ]x[m − k , n − l ]
k =− 1 l =−1

K L
y[ m, n] =   h[k , l ] x[m − k , n − l ]
k =− K l= − L
m +K n+ L
=   h[k , l ] x[ m − k , n − l ]
k =m −K l =n −L
s i tes.google.com/site/ncpdhbkhn 51
 Discrete – Time Signals and
Systems
1. Discrete – Time Signals
2. Discrete – Time Systems
3. Convolution Description of Linear Time – Invariant
Systems
4. Properties of Linear Time – Invariant Systems
5. Analytic Evaluation of Convolution
6. Numerical Computation of Convolution
7. Real – Time Implementation of FIR Filters
8. FIR Spatial Filters
9. Systems Described by Linear Constant –
Coefficient Difference Equations
10. Continuous – Time LTI Systems

s i tes.google.com/site/ncpdhbkhn 52
Systems Described by Linear Constant –
Coefficient Difference Equations (1)
h[ n ] = ba n u[ n ], −1 < a < 1
n n
y[ n] =  x[k ]h[n − k ] =  h[k ]x[n − k ]
k =−∞ k =−∞

= bx[ n] + bax[ n − 1] + ba 2 x[n − 2] + ...

= bx[ n] + a(bx[n − 1] + bax[ n − 2] + ...)
y[ n − 1] = bx [n − 1] + bax[ n − 2] + ...
→ y[ n] = ay[ n − 1] + bx[ n]

b
x[ n ] + y[ n ]

a
z–1
s i tes.google.com/site/ncpdhbkhn 53
Systems Described by Linear Constant –
Coefficient Difference Equations (2)
y[ n] = ay[ n − 1] + bx[ n]
y[ 0] = ay[ −1] + bx[ 0]

y[1] = ay[0] + bx[1] = a 2 y[ −1] + bax[ 0] + bx[1]

y[ 2] = ay[1] + bx[ 2] = a 3 y[ −1] + ba 2 x[ 0] + bax[1] + bx[ 2]

⋮
y[ n] = ay[n − 1] + bx[ n] =
= a n +1 y[−1] + ba n x[0] + ba n −1 x[1] + ... + bx[ n]

h[n ] = ba n u[n ]

→ y[ n] = a n +1 y[n − 1] + h[n ] x[0] + h[ n − 1] x[1] + ... + h[ 0] x[ n ]

s i tes.google.com/site/ncpdhbkhn 54
Systems Described by Linear Constant –
Coefficient Difference Equations (3)
y[ n] = a n +1 y[ n − 1] + h[ n ]x[0] + h[ n − 1] x[1] + ... + h[ 0] x[n ]

If x[n ] = 0 for n ≥ 0 then y zi[n ] = a n+1 y[−1], n ≥ 0

(Zero – input response)
n
If y[ −1] = 0 then yzs [ n] =  h[ k ] x[n − k ]
k =0

(Zero – state response)

n
→ y[ n] = a n +1
y[ n − 1] +  h[ k ] x[ n − k ] = yzi [n ] + yzs [ n]
zero −input k =0
response zero − state
response

s i tes.google.com/site/ncpdhbkhn 55
Systems Described by Linear Constant –
Coefficient Difference Equations (4)
y[ n] = a n +1 y[ n − 1] + h[ n ]x[0] + h[ n − 1] x[1] + ... + h[ 0] x[n ]
x[n ] = u[n]
n +1
n
1 − a
→ y[ n] =  ba k + a n+1 y[−1] = b + a n+1 y[ −1], n ≥ 0
k =0 1− a
a <1
1
→ yss [ n] = lim y[n ] = b , n≥0 (Steady – state response)
n →∞ 1− a
ba n +1
ytr [ n] = y[n ] − y ss [n ] = + a n +1 y[ −1], n ≥ 0
a −1
(Transient response)
lim ytr [n ] = 0
n→∞

b − a n +1 b − a n+1
y[ n] = +b + a y[ −1] =
n +1
+b + a n+1 y[ −1]
1− a 1− a 1− a 1− a
yzi [ n ]
yss [ n ] ytr [ n ] yzs [ n ]

s i tes.google.com/site/ncpdhbkhn 56
 Systems Described by Linear Constant –
Coefficient Difference Equations (5)

6 u[n]
4 y[ n]
2 y tr [ n]
0

-2

-4

-6

-8
0 2 4 6 8 10 12 14 16 18 20

b − a n +1 n +1 b − a n +1
y[ n] = +b + a y[ −1] = +b + a n +1 y[ −1]
1−a 1− a 1− a 1−a y [n] zi
y ss [ n ] y tr [ n ] y zs [ n ]

s i tes.google.com/site/ncpdhbkhn 57
 Systems Described by Linear Constant –
Coefficient Difference Equations (6)

3.5

3
u[ n]
2.5

2 y[ n]
1.5 y tr [ n]
1

0.5

-0.5

-1

-1.5

-2
0 2 4 6 8 10 12 14 16 18 20

b − a n +1 n +1 b − a n +1
y[ n] = +b + a y[ −1] = +b + a n +1 y[ −1]
1−a 1− a 1− a 1−a y [n] zi
y ss [ n ] y tr [ n ] y zs [ n ]

s i tes.google.com/site/ncpdhbkhn 58
 Systems Described by Linear Constant –
Coefficient Difference Equations (7)
n
y[ n] = a n +1
y[ n − 1] +  h[ k ] x[ n − k ]
k =0

x[n ] = e jω0n

 ( ae )
n n
y[ −1] +  a e
n +1 jω 0 ( n − k ) n +1 jω0 n − jω0 k
→ y[ n] = a k
=a y[ −1] + e
k =0 k =0

n +1 − jω ( n +1)
1 − a e 0
n +1
= a y[ −1] + e jω0n
1 − ae − jω0
y zs [ n ] y zs [ n ]
y zi [ n ]
− jω ( n +1 )
−a n +1e 0 1
=a n +1
y[ −1] + e jω0n + e jω0 n

1 − ae− jω0 1 − ae − jω0

y tr [ n ] y ss [ n ]

s i tes.google.com/site/ncpdhbkhn 59
 Systems Described by Linear Constant –
Coefficient Difference Equations (8)
1.5

0.5
Real part

-0.5

-1

0
Input
1.5

Output
1
Imaginary part

0.5

-0.5

-1

0 0

s i tes.google.com/site/ncpdhbkhn 60
 Systems Described by Linear Constant –
Coefficient Difference Equations (9)
N M
y[n ] = −  a k y[n − k ] +  bk x[n − k ]
k =1 k =1

• Linear constant – coefficient difference

equation (LCCDE)
• ak: feedback coefficients
• bk: feedforward coefficients
• If ak & bk are fixed, then the system is time –
invariant
• If they depend on n, then time – varying
• N is the order of the system
s i tes.google.com/site/ncpdhbkhn 61
 Systems Described by Linear Constant –
Coefficient Difference Equations (10)
N M
y[n ] = −  a k y[n − k ] +  bk x[n − k ]
k =1 k =1

• Is the system linear time – invariant?

• Find its impulse response in analytical form
• Given an analytical input x[n], find its
analytical output y[n]
• Is the system stable (given ak & bk)?

s i tes.google.com/site/ncpdhbkhn 62
 Discrete – Time Signals and
Systems
1. Discrete – Time Signals
2. Discrete – Time Systems
3. Convolution Description of Linear Time – Invariant
Systems
4. Properties of Linear Time – Invariant Systems
5. Analytic Evaluation of Convolution
6. Numerical Computation of Convolution
7. Real – Time Implementation of FIR Filters
8. FIR Spatial Filters
9. Systems Described by Linear Constant – Coefficient
Difference Equations
10. Continuous – Time LTI Systems

s i tes.google.com/site/ncpdhbkhn 63
 Continuous – Time LTI Systems
(1)
x (t)
∞ ∞
y[ n ] =  x[k ]h[n − k ] =  h[k ]x[n − k ]
k =−∞ k =−∞
0 t
∞ ∞
y (t ) =  x (τ )h(t − τ )dτ =  h(τ ) x (t − τ )dτ
−∞ −∞ h(t )

0 t

h( t − τ ) x (t − τ )

x(τ ) h (τ )
0 τ 0 τ

s i tes.google.com/site/ncpdhbkhn 64
 Continuous – Time LTI Systems
Ex. 1 (2)
Find the convolution of the two signals?
1
t t h( t )
y ( t ) = h(t )* x ( t ) =  h( t − τ ) x (τ )dτ =  h(τ )τ ( t − τ )dτ
0 0
0 1 2 3 4t

2
x (t)
2
x(τ )
1
h(t − τ ) 0 1 2 3 4t
0 1 2 3 4λ
2
x(τ )
1
0 < t < 1: h(t ) = 1; x(t) = 0 h(t − τ )
0 1 2 3 4λ
h (t )* x ( t ) = 0
0 < t < 1: h(t )* x(t) = 0

s i tes.google.com/site/ncpdhbkhn 65
 Continuous – Time LTI Systems
Ex. 1 (3)
Find the convolution of the two signals?
1
t t h( t )
y ( t ) = h (t )* x ( t ) =  h ( t − τ ) x (τ )dτ =  h (τ )τ ( t − τ )d τ
0 0
0 1 2 3 4t
2
x(τ ) 2
1 x (t)
h( t − τ )
0 1 2 3 4τ
0 1 2 3 4t
2
x(τ ) 0 < t < 1: h (t )* x ( t ) = 0
1
h( t − τ ) 1 < t < 2 : h (t )* x ( t ) = 2( t − 1)
0 1 t 2 3 4τ
1 < t < 2 : h (t ) = 1; x ( t ) = 2
t t
h (t )* x ( t ) =  h ( t − τ ) x (τ ) dτ =  1 × 2 dτ = 2τ τ =1 = 2( t − 1)
t
1 1
s i tes.google.com/site/ncpdhbkhn 66
 Continuous – Time LTI Systems
Ex. 1 (4)
Find the convolution of the two signals?
1
t t h( t )
y ( t ) = h (t )* x ( t ) =  h ( t − τ ) x (τ )dτ =  h (τ )τ ( t − τ )d τ
0 0
0 1 2 3 4t
2
x(τ )
1 2
x (t)
h( t − τ )
0 1 2 3 4λ
0 1 2 3 4t
2
x(τ )
t −1 1 t 0 < t < 1: h (t )* x ( t ) = 0
h( t − τ ) 1 < t < 2 : h (t )* x ( t ) = 2( t − 1)
0 1 2 3 4λ 2 < t < 3: h(t )* x(t) = 2

2 < t < 3: h ( t ) = 1; x ( t ) = 2
t t
h (t )* x ( t ) =  f1 (t − τ ) f 2 (τ ) dτ =  1 × 2 dτ = 2τ τ =t −1 = 2
t
t −1 t −1
s i tes.google.com/site/ncpdhbkhn 67
 Continuous – Time LTI Systems
Ex. 1 (5)
Find the convolution of the two signals?
1
t t h( t )
y ( t ) = h (t )* x ( t ) =  h ( t − τ ) x (τ )dτ =  h (τ )τ ( t − τ )d τ
0 0
0 1 2 3 4t
2
x(τ )
1 2
h( t − τ ) x (t)
0 1 2 3 4λ
0 1 2 3 4t
2
x(τ ) 0 < t < 1: h (t )* x ( t ) = 0
t −1 1 t
h( t − τ ) 1 < t < 2 : h (t )* x ( t ) = 2( t − 1)
0 1 2 3 4λ 2 < t < 3: h(t )* x(t) = 2

3 < t < 4 : h ( t ) = 1; x ( t ) = 2 3< t <4: h(t )* x(t) = 8 − 2t

3 3
h (t )* x ( t ) =  h ( t − τ ) x (τ )d τ =  1 × 2 dτ = 2τ
3
= 8 − 2t
τ =t −1
t −1 t −1
s i tes.google.com/site/ncpdhbkhn 68
 Continuous – Time LTI Systems
Ex. 1 (6)
Find the convolution of the two signals?
1
t t h( t )
y ( t ) = h (t )* x ( t ) =  h ( t − τ ) x (τ )dτ =  h (τ )τ ( t − τ )d τ
0 0
0 1 2 3 4t

2
x (t)
2
x(τ )
1
h(t − τ ) 0 1 2 3 4t
0 1 2 3 4λ
0 < t < 1: h (t )* x ( t ) = 0
t > 4 : h(t ) = 1; x(t ) = 0 1 < t < 2 : h (t )* x ( t ) = 2( t − 1)
2 < t < 3: h(t )* x(t) = 2
h (t )* x ( t ) = 0 3< t <4: h(t )* x(t) = 8 − 2t
t > 4: h( t ) * x ( t ) = 0

s i tes.google.com/site/ncpdhbkhn 69
 Continuous – Time LTI Systems
Ex. 1 (7)
Find the convolution of the two signals?
1
t t h( t )
y ( t ) = h (t )* x ( t ) =  h ( t − τ ) x (τ )dτ =  h (τ )τ ( t − τ )d τ
0 0
0 1 2 3 4t

2
x (t)

0 < t < 1: h(t) * x(t) = 0

0 1 2 3 4t
1 < t < 2 : h(t) * x(t ) = 2(t − 1)
2 < t < 3: h(t) * x(t) = 2 f1(t) * f 2 (t )
3<t <4: h (t )* x ( t ) = 8 − 2t 2
t > 4: h(t) * x(t ) = 0

0 1 2 3 4t

s i tes.google.com/site/ncpdhbkhn 70
 Continuous – Time LTI Systems
Ex. 2 (8)
Find the convolution of the two signals?
t t
1 w(t )
y ( t ) = h (t )* x ( t ) =  h ( t − τ ) x (τ )dτ =  h (τ )τ ( t − τ )d τ
0 0
0 1 2 3 4t

Method 1 2
2e− t
2
2e −τ
w (t )
0 1 2 3 4t
0 1 2 3 4t
2
t τ =t
0 < t < 2 : f (t ) =  2 e −( t −τ ) × 1d τ = 2 eτ −t τ =0 = 2(1 − e − t )
0

0 1 2 3 4t
τ −t τ =2
2
t > 2: f (t ) =  2e −( t −τ ) × 1dτ = 2e τ =0
= 2( e 2 − 1) e − t
0

s i tes.google.com/site/ncpdhbkhn 71
 Continuous – Time LTI Systems
Ex. 2 (9)
Find the convolution of the two signals?
t t
1 w(t )
y ( t ) = h (t )* x ( t ) =  h ( t − τ ) x (τ )dτ =  h (τ )τ ( t − τ )d τ
0 0
0 1 2 3 4t

Method 2 2
2e− t
2
2e − t
w(−τ )
0 1 2 3 4t
0 1 2 3 4t
2
t τ =t
0 < t < 2 : f ( t ) =  1 × 2 e −τ dτ = −2 e −τ τ =0 = 2(1 − e − t )
0

0 1 2 3 4t
−τ τ = t
t
t > 2: f (t ) =  1 × 2 e −τ dτ = −2 e τ = t −2
= 2( e 2 − 1) e − t
t −2

s i tes.google.com/site/ncpdhbkhn 72
 Continuous – Time LTI Systems
(10)
1, n=0 δ ∆ (t)
δ [ n] = 
0, n≠0
−∆ ∆
1, t=0
δ (t ) =  2 2
0, t≠0
0 t
1 / ∆, −∆ /2 < t < ∆/ 2
δ ∆ (t ) = 
0, otherwise
∞
y (t ) =  x(τ )δ ∆ (t − τ )dτ
−∞
x (τ )
x(τ )δ ∆ (t − τ ) ≈ x(t)δ ∆ ( t − τ )
∞
→ y(t ) =  x(τ )δ ∆ (t − τ )dτ
−∞
∞
≈ x(t )  δ ∆ ( t − τ ) dτ
−∞

0
τ = x(t )
t =τ s i tes.google.com/site/ncpdhbkhn 73
Continuous – Time LTI Systems
(11)
∞
y(t ) =  x(τ )δ ∆ ( t − τ ) dτ ≈ x (t ) Aδ ∆ (t )
−∞ A

∞
δ (t ) →  x(τ )δ (t − τ )dτ = x( t )
−∞ 0 t

∞
δ (t ) * x(t ) =  x (τ )δ (t − τ )dτ = x(t )
−∞

x (t )δ ( t − t0 ) = x (t0 )δ ( t − t0 )

∞ ∞
 x( t )δ (t − t0 ) dt = x(t0 )  δ ( t − t0 )dt = x( t0 )
−∞ −∞

s i tes.google.com/site/ncpdhbkhn 74