455DTSigSys Handouts

Discrete-Time Signals and Systems
Reference:
Sections 2.1-2.5 of
Professor Deepa Kundur
John G. Proakis and Dimitris G. Manolakis, Digital Signal Processing:

Principles, Algorithms, and Applications, 4th edition, 2007.
University of Toronto
Professor Deepa Kundur (University of Toronto)
1 / 63
2 / 63
Elementary Discrete-Time Signals

1. unit sample sequence (a.k.a. Kronecker delta function):

1, for n = 0
(n) =
0, for n 6= 0
2. unit step signal:
2.1 Discrete-Time Signals
1,
0,
n, for n 0
0, for n < 0
u(n) =
for n 0
for n < 0
3. unit ramp signal:

ur (n) =
Note:
(n)
= u(n) u(n 1) = ur (n + 1) 2ur (n) + ur (n 1)
u(n) = ur (n + 1) ur (n)
3 / 63
4 / 63
Signal Symmetry
Signal Symmetry
Even signal:
x(n) = x(n)
Even signal component: xe (n) = 12 [x(n) + x(n)]
Odd signal: x(n) = x(n)
Odd signal component: xo (n) = 12 [x(n) x(n)]

x(n)
x(n)
x(n)
-7
-6
-5
-4
-3
-2
-1
-7
-6
-5
-4
-3
-2
-1
-7
-6
-5
-4
-3
-2
-1
Note: x(n) = xe (n) + xo (n)
5 / 63
Simple Manipulation of Discrete-Time Signals
x(n)
I
0.5
-6 -5 -4 -3 -2 -1 0 1
-0.5
2 3 4
5 6
7 8
Transformation of independent variable:

I
7 / 63
Question: what if 6 Z?
Additional, multiplication and scaling:

I
x(n)
Question: what if k 6 Z?
time scale: n n, Z
I
time shift: n n k, k Z
I
-1
6 / 63
Signal Symmetry
1
amplitude scaling: y (n) = Ax(n),

< n <
sum: y (n) = x1 (n) + x2 (n),
< n <
product: y (n) = x1 (n)x2 (n),
< n <
8 / 63
Simple Manipulation of Discrete-Time Signals
Simple Manipulation of Discrete-Time Signals I
Find x(n) x(n + 1).
Find x( 32 n + 1).
x(n)
3
2
1
-3 -2 -1 0 1
2 3 4
-1
5 6
15
7 8 9 10
20
-1
-2
-2
-3
x(n)
2 3 4
5 6
9 / 63
15
7 8 9 10
20
Simple -1Manipulation of Discrete-Time Signals

-1
-1
Graph of-2x( 32 n + 1).
0 if
3n
2
0 if
3n
2
5
2
4
11
2
17
2
10
23
2
13
29
2
16
35
2
19
> 19
x( 3n
+ 1)
2
+ 1 is an integer; undefined otherwise
undefined
x(1) = 1
undefined
x(4) = 2
undefined
x(7) = 3
undefined
x(10) = 2
undefined
x(13) = 1
undefined
x(16) = 1
undefined
x(19) = 2
+ 1 is an integer; undefined otherwise
10 / 63
-1
-1
-x(n+1)
3
2
1
-3 -2 -1 0 1
-1
+1
< 12
12
1
-3 -2 -1 0 1
3n
2
x(n)-x(n+1)
-3
n
< 1
-1
0
1
2
3
4
5
6
7
8
9
10
11
12
> 12
2 3 4
5 6
-2
-3
This signal is undefined for values of n

that are not even integers and zero for
even integers not shown on this sketch.
7 8 9
10
11
12
13 14
2.2 Discrete-Time Systems

n
-1
-2
11 / 63
12 / 63
Terminology: Implication
Terminology: Equivalence
If A then B
Shorthand: A = B
If A then B
Shorthand: A = B
and
Example 1:
it is snowing = it is at or below freezing temperature
If B then A
Example 2:
5.2 = is positive
Note: For both examples above, B 6= A
Shorthand: B = A
can be rewritten as
Shorthand: A B
A if and only if B
We can also say:
13 / 63
A is EQUIVALENT to B
A=B
x(n)
Discrete-time
System
y(n)
Classification of Discrete-Time Systems

Why is this so important?
output/
response
mathematical techniques developed to analyze systems are often

contingent upon the general characteristics of the systems being
considered
For a system to possess a given property, the property must hold

for every possible input to the system.
Discrete-time
signal
Discrete-time
signal
Input-output description (exact structure of system is unknown

or ignored):
y (n) = T [x(n)]
14 / 63
Terminology: Input-Output Description

input/
excitation
black box representation:
to disprove a property, need a single counter-example

to prove a property, need to prove for the general case
x(n) y (n)
15 / 63
16 / 63
Classification of Discrete-Time Systems
Static vs. Dynamic
Common System Properties:

static
vs.
dynamic
time-invariant
vs.
time-variant
linear
vs.
nonlinear
causal
vs.
non-causal
stable
vs.
unstable systems
..
.
Static system (a.k.a. memoryless): the output at time n

depends only on the input sample at time n; otherwise the
system is said to be dynamic
a system is static iff (if and only if)

y (n) = T [x(n), n]
for every time instant n.
..
.
17 / 63
Static vs. Dynamic

Examples: memoryless or not?
Consider the general system:
y (n)
= T [x(n N), x(n N + 1), , x(n 1), x(n), x(n + 1),

, x(n + M 1), x(n + M)],
N, M > 0
I
I
For N = M = 0, y (n) = T [x(n)], the system is static.
For 0 < N, M < , the system is said to be dynamic with finite

memory of duration N + M + 1.
For either N and/or M equal to infinite, the system is said to

have infinite memory.
18 / 63
Static vs. Dynamic

I
19 / 63
I
I
I
y (n) = A x(n), A 6= 0
y (n) = A x(n) + B, A, B, 6= 0
y (n) = x(n) cos( 25

(n 5))
y (n) = x(n)
y (n) = x(n + 1)
1
y (n) = 1x(n+2)
y (n) = e 3x(n)
P
y (n) = nk= x(k)
Ans: Y, Y, Y, N, N, N, Y, N
20 / 63
Time-invariant vs. Time-variant Systems
Time-invariant vs. Time-variant Systems

Examples: time-invariant or not?
Time-invariant system: input-output characteristics do not

change with time
I
I
I
a system is time-invariant iff
I
T
x(n) y (n) = x(n k) y (n k)
I
I
for every input x(n) and every time shift k.
y (n) = A x(n), A 6= 0
y (n) = A x(n) + B, A, B, 6= 0
y (n) = x(n) cos( 25

n)
y (n) = x(n)
y (n) = x(n + 1)
1
y (n) = 1x(n+2)
y (n) = e 3x(n)
P
y (n) = nk= x(k)
Ans: Y, Y, N, N, Y, Y, Y, Y
21 / 63
22 / 63
Linear vs. Nonlinear Systems
Linear Systems: Homogeneity

A system is linear iff
Linear system: obeys superposition principle
a system is linear iff
T [a1 x1 (n) + a2 x2 (n)] = a1 T [x1 (n)] + a2 T [x2 (n)]
Homogeneity: Let a2 = 0.
T [a1 x1 (n)] = a1 T [x1 (n)]

for any arbitrary input sequences x1 (n) and x2 (n), and any
arbitrary constants a1 and a2
x(n) y (n) = a1 x(n) a1 y (n)

for any constant a1 .
23 / 63
24 / 63
Linear Systems: Additivity
Linear Systems: Additivity
A system is linear iff

Therefore:
Linearity
Additivity: Let a1 = a2 = 1.
Homogeneity
Additivity
T [x1 (n) + x2 (n)] = T [x1 (n)] + T [x2 (n)]
x1 (n) y1 (n)
T
x2 (n) y2 (n)
Need both!
If a system is not homogeneous, it is not linear.
If a system is not additive, it is not linear.
= x1 (n) + x2 (n) y1 (n) + y2 (n)
for any input sequences x1 (n) and x2 (n).

25 / 63
26 / 63
Linear vs. Nonlinear Systems
Causal vs. Noncausal Systems
Examples: linear or not?

I
I
I
I
I
I
I
I
y (n) = A x(n), A 6= 0
y (n) = A x(n) + B, A, B, 6= 0
y (n) = x(n) cos( 25

n)
y (n) = x(n)
y (n) = x(n + 1)
1
y (n) = 1x(n+2)
Causal system: output of system at any time n depends only on

present and past inputs
a system is causal iff

y (n) = T [x(n), x(n 1), x(n 2), . . .]
y (n) = e 3x(n)
P
y (n) = nk= x(k)
for all n
Ans: Y, N, Y, Y, Y, N, N, Y
27 / 63
28 / 63
Causal vs. Noncausal Systems
Stable vs. Unstable Systems
Examples: causal or not?

I
I
I
I
I
I
I
I
y (n) = A x(n), A 6= 0
y (n) = A x(n) + B, A, B, 6= 0
y (n) = x(n) cos( 25

(n + 1))
y (n) = x(n)
y (n) = x(n + 1)
1
y (n) = 1x(n+2)
Bounded Input-Bounded output (BIBO) Stable: every bounded

input produces a bounded output
a system is BIBO stable iff

|x(n)| Mx < = |y (n)| My <
y (n) = e 3x(n)
P
y (n) = nk= x(k)
for all n.
Ans: Y, Y, Y, N, N, N, Y, Y
29 / 63
Stable vs. Unstable Systems

Examples: stable or not?
x[n]
I
I
|x[n]|
|x[n]|
I
n
I
I
x[n]
x[n]
I
I
BOUNDED SIGNAL
30 / 63
Discrete-Time Bounded Signals

x[n]
y (n) = A x(n), A 6= 0
y (n) = A x(n) + B, A, B, 6= 0
y (n) = x(n) cos( 25

n)
y (n) = x(n)
y (n) = x(n + 1)
1
y (n) = 1x(n+2)
y (n) = e 3x(n)
P
y (n) = nk= x(k)
UNBOUNDED SIGNAL n
Ans: Y, Y, Y, Y, Y, N, Y, N
31 / 63
32 / 63
Final Remarks
For a system to possess a given property, the property must hold

for every possible input and parameter of the system.
to disprove a property, need a single counter-example
to prove a property, need to prove for the general case
2.3 Analysis of Dst-Time Linear

Time-Invaraint Systems
33 / 63
34 / 63
The Convolution Sum
The Convolution Sum

Let the response of a linear time-invariant (LTI) system to the unit
sample input (n) be h(n).
T
(n) h(n)
Recall:
x(n) =
(n k) h(n k)
x(k)(n k)
(n k) h(n k)
k=
x(k) (n k) x(k) h(n k)
X
X
T
x(k)(n k)
x(k)h(n k)
k=
k=
T
x(n) y (n)
35 / 63
36 / 63
The Convolution Sum
Properties of Convolution
Associative and Commutative Laws:
x(n) h(n) = h(n) x(n)
[x(n) h1 (n)] h2 (n) = x(n) [h1 (n) h2 (n)]
Therefore,
y (n) =
x(k)h(n k) = x(n) h(n)
x(n)
h1(n)
k=
y(n)
h1(n) * h2(n) = h2(n) * h1(n)
for any LTI system.

x(n)
h2(n)
37 / 63
h2(n)
h1(n)
y(n)
38 / 63
Properties of Convolution
Causality and Convolution
Distributive Law:
For a causal system, y (n) only depends on present and past inputs
values. Therefore, for a causal system, we have:
x(n) [h1 (n) + h2 (n)] = x(n) h1 (n) + x(n) h2 (n)

y (n) =
h1(n) + h2(n)
h(k)x(n k)
k=
h1(n)
x(n)
y(n)
h2(n)
1
X
h(k)x(n k) +
k=
h(k)x(n k)
k=0
h(k)x(n k)
k=0
where h(n) = 0 for n < 0 to ensure causality.
39 / 63
40 / 63
Finite vs. Infinite Impulse Response

For causal LTI systems, h(n) = 0 for n < 0.
Finite impulse response (FIR):
y (n) =
2.4 Discrete-time Systems

Described by Difference Equations
M1
X
h(k)x(n k)
k=0
Infinite impulse response (IIR):

y (n) =
h(k)x(n k)
k=0
How would one realize these systems?

41 / 63
There is a practical and computationally efficient means of

implementing all FIR and a family of IIR systems that makes use of
...
. . . difference equations.
Finite impulse response (FIR):

)
h(k)x(n k)
42 / 63
System Realization
Implementation: Two classes
y (n) =
Finite vs. Infinite Impulse Response
M1
X
All systems
nonrecursive systems
All LTI systems
k=0
Infinite impulse response (IIR):

y (n) =
)
h(k)x(n k)
recursive systems
k=0
43 / 63
44 / 63
System Realization
System Realization

...

...
All systems
All systems
All LTI systems
All LTI systems
Systems
Described
by LCCDEs
All LTI FIR systems
All LTI IIR systems
All LTI FIR systems
45 / 63
All LTI IIR systems
46 / 63
System Realization
General expression for Nth-order LCCDE:
N
M
X
X
ak y (nk) =
bk x(nk)
k=0
a0 , 1
2.5 Implementation of
Discrete-time Systems
k=0
Initial conditions: y (1), y (2), y (3), . . . , y (N).
Need: (1) constant scale, (2) addition, (3) delay elements.
47 / 63
48 / 63
Building Block Elements
Adder:
FIR System Realization
Unit delay:
Finite Impulse Response Systems and Nonrecursive Implementation

Constant multiplier:
Unit advance:
Signal multiplier:
49 / 63
FIR System Realization: Example
Consider a 5-point local averager:
n
1 X
y (n) =
x(k) n = 0, 1, 2, . . .
5 k=n4
50 / 63

I

n
1 X
y (n) =
x(k) n = 0, 1, 2, . . .
5 k=n4
The impulse response is given by:
h[n]
n
1 X
h(n) =
(k)
5 k=n4
1/5
1
1
1
(n 4) + (n 3) + (n 2) +
5
5
5
1
1
(n 1) + (n)
5
5
1 2 3 4
n
Indeed FIR!
51 / 63
52 / 63

n
1 X
y (n) =
x(k) n = 0, 1, 2, . . .
5 k=n4
y (n)
n
n
X
1 X
1
x(k) =
x(k)
5
5
k=n4
y (n) =
k=n4
1
1
1
x(n 4) + x(n 3) + x(n 2) +
5
5
5
1
1
x(n 1) + x(n)
5
5
Memory requirements stay constant; only need to store 5 values

(4 last + present).
fixed number of adders required
1/5
1/5
+
53 / 63
y (n) =
Adder:
54 / 63
...
Unit delay:
Infinite Impulse Response Systems and Recursive Implementation
+
Unit advance:
Requires:
I M multiplications
I M 1 additions
I M 1 memory elements
bk x(n k)
...
+
1/5
FIR System Realization
k=0
1/5
FIR System Realization: General

M1
X
1/5
Signal multiplier:
55 / 63
56 / 63
IIR System Realization: Example

I
Consider an accumulator:
y (n) =
n
X
x(k) n = 0, 1, 2, . . .
for y (1) = 0.
y (n) =
k=0
n
X
x(k) n = 0, 1, 2, . . .
for y (1) = 0.
k=0
I
The impulse response is given by:

h(n) =
Consider an accumulator:
n
X
IIR memory requirements seem to grow with increasing n!
(k) = (n) + (n 1) + (n 2) +
k=0

=
1 n0
0 n<0
h[n]
1
...
0 1 2 3 4
n
57 / 63
y (n) =
k=0
n1
X
y (n) =
N
X
k=1
bk x(n k)
k=0
M
X
bk x(n k)
nonrecursive
| {z }
k=0
input 1
N
X
y (n)
=
ak y (n k) + v (n) recursive
|{z}
|{z}
k=1
output 2
input 2
v (n)
=
|{z}
output 1
k=0
recursive implementation
ak y (n k) +
M
X
is equivalent to the cascade of the following systems:
x(k) + x(n)
= y (n 1) + x(n)
y (n) = y (n 1) + x(n)
58 / 63
Direct Form I vs. Direct Form II Realizations
x(k)
n
X
59 / 63
60 / 63
Direct Form I IIR Filter Implementation

v(n)
LTI All-zero system
LTI All-pole system
LTI All-pole system
Requires: M + N + 1 multiplications, M + N additions, M + N memory locations

61 / 63
LTI All-zero system
Requires: M + N + 1 multiplications, M + N additions, M + N memory locations

Direct Form II IIR Filter Implementation

+
Unit delay:
Unit advance:
Unit advance:
...
...
Signal multiplier:
...
...
For N>M
Requires: M + N + 1 multiplications, M + N additions, max(M, N) memory

locations
Unit delay:
Constant multiplier: +
Signal multiplier:
Adder:
63 / 63
Adder:
...
...
...
...
...
...
...
...
Direct Form II IIR Filter Implementation
62 / 63

455DTSigSys Handouts

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

455DTSigSys Handouts

Uploaded by

Copyright:

Available Formats

Discrete-Time Signals and Systems

Discrete-Time Signals and Systems

Discrete-Time Signals and Systems

Professor Deepa Kundur

John G. Proakis and Dimitris G. Manolakis, Digital Signal Processing:

Professor Deepa Kundur (University of Toronto)

Discrete-Time Signals and Systems

Professor Deepa Kundur (University of Toronto)

Discrete-Time Signals and Systems

Discrete-Time Signals and Systems

Discrete-Time Signals and Systems

Elementary Discrete-Time Signals

2. unit step signal:

2.1 Discrete-Time Signals

3. unit ramp signal:

= u(n) u(n 1) = ur (n + 1) 2ur (n) + ur (n 1)

Discrete-Time Signals and Systems

Professor Deepa Kundur (University of Toronto)

Discrete-Time Signals and Systems

Discrete-Time Signals and Systems

Discrete-Time Signals and Systems

Odd signal: x(n) = x(n)

Odd signal component: xo (n) = 12 [x(n) x(n)]

Professor Deepa Kundur (University of Toronto)

Discrete-Time Signals and Systems

Note: x(n) = xe (n) + xo (n)

Professor Deepa Kundur (University of Toronto)

Discrete-Time Signals and Systems

Simple Manipulation of Discrete-Time Signals

Transformation of independent variable:

Discrete-Time Signals and Systems

Additional, multiplication and scaling:

Professor Deepa Kundur (University of Toronto)

Discrete-Time Signals and Systems

Discrete-Time Signals and Systems

amplitude scaling: y (n) = Ax(n),

Professor Deepa Kundur (University of Toronto)

Discrete-Time Signals and Systems

Discrete-Time Signals and Systems

Discrete-Time Signals and Systems

Simple Manipulation of Discrete-Time Signals

Simple Manipulation of Discrete-Time Signals I

Find x(n) x(n + 1).

Professor Deepa Kundur (University of Toronto)

Simple -1Manipulation of Discrete-Time Signals

Graph of-2x( 32 n + 1).

Professor Deepa Kundur (University of Toronto)

Discrete-Time Signals and Systems

Discrete-Time Signals and Systems

Discrete-Time Signals and Systems

Discrete-Time Signals and Systems

Professor Deepa Kundur (University of Toronto)

This signal is undefined for values of n

2.2 Discrete-Time Systems

Discrete-Time Signals and Systems

Professor Deepa Kundur (University of Toronto)

Discrete-Time Signals and Systems

Discrete-Time Signals and Systems

Discrete-Time Signals and Systems

Professor Deepa Kundur (University of Toronto)

Discrete-Time Signals and Systems

Professor Deepa Kundur (University of Toronto)

Discrete-Time Signals and Systems

Classification of Discrete-Time Systems