EE2010/IM2004 Signals and Systems

Academic Year: Semester 1

Part I: Weeks 1-5
Lecturer: Prof. Ma Kai-Kuang
Office: S2-B2C-83
Tel: 6790-6366
Email: [email protected]

Notes and the attached tutorial set are from Assoc/Prof Teh Kah Chan.
Refer to NTUlearn regularly for important announcement, including rules
and assessment criteria.

Subject Outline
Signals and Systems
Classication of Signals
Elementary and Singularity Signals
Operations on Signals
Properties of Systems
Linear Time-Invariant (LTI) Systems
Continuous-Time and Discrete-Time LTI Systems
Convolution
LTI System Properties
Correlation Functions

Textbook
1. M. J. Roberts, Fundamentals of Signals and Systems,
McGraw-Hill, International Edition, 2008. (TK5102.9.R646F)
References
1. M. J. Roberts, Signals and Systems, McGraw-Hill, International
Edition, 2003. (TK5102.9.R63)
2. A. V. Oppenheim and A. S. Willsky, Signals and Systems,
Prentice-Hall, 2nd Edition, 1997. (QA402.P62)
3. S. Haykin and B. V. Veen, Signals and Systems, Wiley, 2nd
Edition, 2003. (TK5102.5.H419)
4. S. S. Soliman and M. D. Srinath, Continuous and Discrete Signals
and Systems, Prentice-Hall, 2nd Edition, 1998. (TK5102.9.S686)
5. B. P. Lathi, Linear Systems and Signals, Oxford University Press,
1st Edition, 2002. (TK5102.5.L352)
3

Overviews of Signals and Systems

Input
Message

Input
Transducer

Input
Signal
Transmitter

Transmitted
Signal

Channel

Output
Message

Output
Transducer

Output
Signal

Receiver

Received
Signal

Figure 1: A typical signal and system example

4

Distortion
and Noise

Classication of Signals
Continuous-Time vs Discrete-Time Signal
Continuous-Value vs Discrete-Value Signal
Deterministic vs Random Signal
Even vs Odd Signal
Periodic vs Aperiodic Signal
Energy-Type vs Power-Type Signal

Continuous-Time vs Discrete-Time Signal

Continuous-Time (CT) Signal: A signal x(t) that is specied for
all value of time t
Discrete-Time (DT) Signal: A signal yn] that is specied only
for integer value of n
y [n]

x (t )

3
0

-1

4
5

Figure 2: Continuous-time vs discrete-time signal

6

7 n

Example 1: Sketch the waveforms of the CT signal x(t) = t and DT

signal xn] = n, respectively.
x (t )

x [ n]

5
-3 -2
0

-5

-5

3
2
-1 1
-1 0 1
-2
-3

Figure 3: Examples of CT and DT signals

Continuous-Value vs Discrete-Value Signal

Continuous-Value Signal: A signal x(t) whose amplitude can take
on any value
Discrete-Value Signal: A signal y(t) whose amplitude can take on
only a nite number of values
y (t)

x( t )

2T

Figure 4: Continuous-value vs discrete-value signal

8

Example 2: Sketch the waveforms of the continuous-value signal

x(t) = A sin(2 f0 t) and discrete-value signal yn] = (;1)n ,
respectively.
x (t)

y[n]

1
-3

0
-A

-2

-1
0

-1

Figure 5: Examples of continuous-value vs discrete-value signals

9

Deterministic vs Random Signal

Deterministic Signal: A signal x(t) that can be mathematically
modeled explicitly as a function of time, i.e., x(t) = A sin(2 f0 t)
Random Signal: A signal y(t) that is known only in terms of
probabilistic description, i.e., noise
x (t )

y (t )

A
0
-A

Figure 6: Deterministic vs random signal

10

Even vs Odd Signal

Even Signal: A signal xe (t) that satises the condition
xe (t) = xe (;t)
Odd Signal: A signal xo (t) that satises the condition
xo (t) = ;xo (;t)
x o( t )

x e( t )

Figure 7: Even vs odd signal

11

Any deterministic signal x(t) can be decomposed into sum of an

even and an odd signal

x(t) = xe (t) + xo (t)

where

xe (t) = 12 x(t) + x(;t)]

and

xo (t) = 12 x(t) ; x(;t)]

12

The product of two even signals is an even signal

The product of two odd signals is an even signal
The product of an even signal and an odd signal is an odd signal
Note that
T0

;T0

xe (t)dt = 2

T0

Z
0

xe (t)dt

and
Z

T0

;T0

xo (t)dt = 0

13

Example 3: Show that the signal x(t) = A sin(2 f0 t) is an odd signal.

Since

x(;t) = A sin2 f0 (;t)]

= ;A sin(2 f0 t)
= ;x(t)
hence, x(t) is an odd signal.
x (t )
A
-A

Figure 8: An odd signal example

14

Example 4: Find the even and odd components of the signal

x(t) = cos(t) + sin(t) cos(t).
The even component of x(t) is
xe (t)

1
=
[x(t) + x(t)]
2
1
=
[cos(t) + sin(t) cos(t) + cos(t) + sin(t) cos(t)]
2
= cos(t)

The odd component of x(t) is

1
[x(t) x(t)]
xo (t) =
2
1
=
[cos(t) + sin(t) cos(t) cos(t) sin(t) cos(t)]
2
= sin(t) cos(t)
15

Example 5: Evaluate

 T0

3
3
x(t)dt
where
x(t)
=
t
cos
(10t).
T0

Since
x(t) = (t)3 cos3 [10(t)]
=

t3 cos3 (10t)

x(t)

hence, x(t) is an odd signal. Thus,

 T0
x(t)dt
T0

16

Periodic vs Aperiodic Signal

Periodic Signal: A signal x(t) with a constant period 0 < T0 < 1
that

x(t) = x(t + T0 )

;1 < t < 1

For a discrete-time signal, the constant period is an integer

0 < K0 < 1 that

xn] = xn + K0 ]

;1 < n < 1

Aperiodic Signal: A signal y(t) or yn] that does not satisfy the
above equation
17

y (t )

x (t )

...

...
T0

2T0

T0

x [ n]

y [n]

...
0

...
1 2
K0=2

Figure 9: Periodic vs aperiodic signal

18

Energy-Type vs Power-Type Signal

Energy-Type Signal

{ A signal x(t) or xn] that has nite energy, i.e., 0 < Ex < 1,
where

CT signal:

Ex =

DT signal:

Ex =

;1
1
X

jx(t)j

n=;1

dt

jxn]j

Power-Type Signal

{ A signal x(t) or xn] that has nite power, i.e., 0 < Px < 1,
where

19

CT signal:
DT signal:

Z T=2
1
2
Px = Tlim
jx(t)j dt
!1 T

;T=2

K
X
1
jxn]j2
Px = Klim
!1 2K + 1 n=;K

{ Note that if x(t) or xn] is a periodic signal with period T0 or

K0 , respectively, then
CT signal:
DT signal:

Z t1 +T0

Px = T1
0 t1

jx(t)j

K0 ;1
1 k+X
2
jxn]j

Px = K
0 n=k

with any real value of t1 and any integer value of k.

20

dt

Example 6: Determine the energy and power of the periodic signal

x(t) = A cos(2 f0 t):

Ex =
=
1
Px = Tlim
!1 T
1

;1
1

;1
T=2

;T=2

T0 =2

= T
0 ;T0 =2

jx(t)j

dt

jA cos(2

f0 t)j2 dt =

1
jx(t)j dt =
T
2

T0 =2

;T0 =2

jx(t)j

dt

2
A
jA cos(2 f0 t)j dt =
2
2

Hence, x(t) is a power-type signal. In general, power-type signals are

periodic signals.
21

Example 7: Determine the energy and power of the signal

y(t) = exp(;jtj):

Ey =
=

;1
1

jy (t)j

dt

jexp(;jtj)j

;1Z

= 2

dt

exp(;2t)dt = 1

T=2
1
2
jy (t)j dt
Py = Tlim
!1 T ;T=2
1
= lim  E = 0
Z

T !1 T

Hence, y(t) is an energy-type signal. In general, energy-type signals

are aperiodic signals.
22

Example 8: Determine the energy and power of the discrete-time

periodic signal xn] = A sin(2 n=4):

Ex =
=

1
X
n=;1

1
X

n=;1

jxn]j

jA sin(2

n=4)j2 =

k+X
K0 ;1
3
X
1
1
jxn]j2 =
jA sin(2 n=4)j2
Px = K
4 n=0
0 n=k
2
2
A
A
2
2
2
2

= 4  0 + 1 + 0 + (;1) = 2

Hence, xn] is a power-type signal.

23

Example 9: A simplied transmitter model of a digital

communication system is shown in Figure 10, determine the
classications of each signal.
x (t) = Acos (2 f0 t )

Ideal
Sampling

x [ n]

x [ n]
Quantization

y [ n]

Discrete Noise
w [ n]

Figure 10: Transmitter model of a digital communication system

24

The waveforms of various signals are shown in Figure 11.

x(t) is a continuous-time, continuous-value, deterministic, even,

periodic, and power-type signal

xn] = x(nTs ) is a discrete-time, continuous-value, deterministic,

even, periodic, and power-type signal

x^n] is a discrete-time, discrete-value, deterministic, even,

periodic, and power-type signal

wn] is a discrete-time, continuous-value, random, and aperiodic

signal

yn] is a discrete-time, continuous-value, random, and aperiodic

signal

25

x( t )

x [ n]

x [ n]

y [ n]

Figure 11: Waveforms of various signals for Example 9

26

Elementary and Singularity Signals

Exponential signal

x(t) = A exp (at)

{ x(t) is growing if a > 0
{ x(t) is decaying if a < 0
exp ( at )

exp ( at )
1

a >0

a< 0

1
0

Figure 12: Exponential signal

27

Sinusoidal signal
x(t) = A cos (2 f0 t + ) or A sin (2 f0 t + )
where A is the amplitude, f0 is the frequency in Hertz, and  is
the phase angle in radians
{ A sinusoidal signal is periodic with period T0 = 1=f0

...

A cos ( 2 f0 t )

...

T0

-A

Figure 13: CT sinusoidal signal

28

The discrete time version of the sinusoidal signal is





xn] = A cos 2 n +  or A sin 2 n + 

K0
K0
where A is the amplitude, K0 is a positive integer dened as the
fundamental period, and  is the phase angle in radians
2n)
n
x
(
[
]
=
A
cos
A
8

... .

. . .
8

...
.

-A

Figure 14: DT sinusoidal signal

29

Complex exponential signal

A exp(j 2 f0 t) = A cos (2 f0 t) + jA sin (2 f0 t)

{ The magnitude of complex exponential signal is given by

jA exp(j 2

f0 t)j = A

{ The sinusoidal signal can be expressed as

A cos (2 f0 t + ) =

< fA exp(j 2

f0 t) exp(j)g

A sin (2 f0 t + ) =

= fA exp(j 2

f0 t) exp(j)g

and

30

Example 10: Sketch the function x(t) = 5 exp(;at)  cos (2 10t)

for t > 0. Assume that a > 0.
x (t )
5

0.1

0.2

0.3

-5

Figure 15: An exponentially damped sinusoidal signal

31

The DT unit impulse (or Dirac Delta) function n] is dened as

n] =

8
<
:

[n]

A [n 3]
A

. ....

-1 0

1 n = 0
0 n 6= 0:

.... .

-1 0

Figure 16: DT impulse functions

32

The CT unit impulse (or Dirac Delta) function (t) is dened as

(t) =

8
< 1
:

(t )

A (t T0 )

 t = 0
0 t 6= 0:

Figure 17: CT impulse functions

33

T0

Properties of the CT impulse function

{ Property 1
Z

;1

(t)dt = 1

{ Property 2
x(t)   (t ; T0 ) = x(T0 )   (t ; T0)

{ Property 3
Z

1
;1

x(t)   (t ; T0 ) dt = x(T0 )

34

The CT unit step function u(t) is dened as

u(t) =

8
<
:

1 t  0
0 t < 0:

u (t )
1

Figure 18: A CT unit step function

35

The DT unit step function un] is dened as

un] =

8
<
:

1 n  0
0 n < 0:

u [ n]

...

.....

...
1

Figure 19: A DT unit step function

36

The CT signum function sgn(t) is dened as

8
>
>
<

sgn(t) =

>
>
:

1
0

t > 0
t = 0
;1 t < 0:

sgn(t )
1
0

t
-1

Figure 20: A CT signum function

37

The DT signum function sgn(t) is dened as

8
>
>
<

sgnn] =

>
>
:

1 n > 0
0 n = 0
;1 n < 0:

sgn [n]

...

1
-6 -5 -4 -3 -2 -1

...

0
-1

Figure 21: A DT signum function

38

;t

The CT unit rectangular function rect T is dened as



rect Tt

8
<
:

1 jtj  T=2
0 otherwise.

rect (Tt )
1

T
2

T
2

Figure 22: A CT unit rectangular function

39



n
The DT unit rectangular function rect K (assume that K is

even) is dened as

i
n
rect K =
h

8
<
:

1 jnj  K=2
0 otherwise.

rect[Kn ]
1

...

...

...
K

-1 0

...

Figure 23: A DT unit rectangular function

40

The sinc function sinc(t) is dened as

sin(
t
)
sinc(t) =

sinc (t )
1

-4

-3

-2

-1

Figure 24: A sinc function

41

4 t

Example 11: The function x(t) = 5  sinc(t) is sampled at every

Ts = 0:5 second interval to produce the sampled signal xs (t), sketch
the waveforms for x(t) and xs (t), respectively.

xs (t) =
=
=

1
X
n=;1

1
X

n=;1

1
X

n=;1

x(t)  (t ; nTs )

x(nTs )  (t ; nTs )
5  sinc(nTs )  (t ; nTs )

42

x (t ) = 5sinc(t)
5

-4

-3

-2

-1

xs (t )
5

. . . .

-4

-3

-2

-1

. . . .
1

Figure 25: Waveforms for x(t) and xs (t)

43

4 t

Operations on Signals
Amplitude scaling: The operation Ax(t) (or Axn]) is to multiply
the amplitude of x(t) (or xn]) by an amount A
x (t)

2 x (t)

1.5 x (t)

4
3
2
2
0

2 t

2 t

-2
-3
-4

Figure 26: Amplitude scaling of signals

44

Time shifting: The operation x(t ; T ) (or xn ; K ]) is to shift

x(t) (or xn]) by an amount T (or K )
x( t)

2
0

0 0.5

2.5

-1 0

x[n 1]

x [ n]

x [ n +1]

3
0

x ( t +1)

x ( t 0.5)

3
1

-1

Figure 27: Time shifting of signals

45

;t

Example 12: Show that rect T = u

1
T
2

;

T
T
t+ ;u t; .
2

u( t + T2 )
t

0
1

u( t T2 )
T
2

rect (Tt ) = u( t + T2 ) u(t T2 )

1
T
2

T
2

Figure 28: Example on time shifting operation

46

CT time scaling: The operation x(t=a) is to scale x(t) by a

{ It expands the function horizontally by a factor jaj

{ If a < 0, the function will be also time inverted
x( t / 2)

x (t )
A
0

x ( 2t)

A
2

A
4 t

Figure 29: CT time scaling of signals

47

-1 0

DT time scaling: xKn] or xn=K ] where K is an integer

{ xKn] : Time compression or decimation

{ xn=K ] : Time expansion
x [ 2n]

x[n]
3

6
5
4

12
-1 0 1 2 3 4

x[n/2]

4
2

-2 -1 0

. ... .

1 2
-2 0

Figure 30: DT time scaling of signals

48

5
6

8 n

;
t

4
;

t;2
2

Example 13: If x(t) = 0:5 rect

the waveform y(t) = ;2x
.

as shown in Figure 31, sketch

x ( t)
0.5

-2

Figure 31: Example of operations on signals

49

x (t)
0.5
-2

2 x ( t /2)
4

-4
0

-1

y( t) = 2 x( t 2 )
2

-2
0

-1

Figure 32: Example of operations on signals

50

Continuous-Time and Discrete-Time Systems

A system refers to any physical device (i.e., communication
channels, lters) that produces an output signal y(t) in response
to an input signal x(t)

x (t)

y (t)

x [n]

y [n]

Figure 33: Block diagram representation of a system

51

Properties of Systems
Stability

{ A system is said to be bounded-input bounded-output (BIBO)

stable if and only if every bounded input (i.e., jx(t)j < 1 for
all t, or jxn]j < 1 for all n) results in bounded output
{ An example of a BIBO stable system

yn] = rn xn]un] jrj < 1

{ An example of a BIBO unstable system

yn] = rn xn]un] jrj > 1
52

Memory

{ A system is said to possess memory if its output signal depends

on past or future values of the input signal
{ An example of a system with memory

yn] = xn] + xn ; 1] + xn ; 2]

{ A system is memoryless if its output signal depends only on the

present value of the input signal

{ An example of a memoryless system

y(t) = x2 (t)

53

Causality

{ A system is causal if the present value of the output signal

depends only on the present or past values of the input signal

{ An example of a causal system
1
yn] = 3 (xn] + xn ; 1] + xn ; 2])

{ A system is noncausal if the present value of the output signal

depends on the future values of the input signal

{ A noncausal system is not physically realizable in real time
{ An example of a noncausal system
yn] = 13 (xn + 1] + xn] + xn ; 1])
54

Linearity

{ A system is linear if the principle of superposition holds, i.e., if

input signal is x3 (t) = a1 x1 (t) + a2 x2 (t), then the output signal

is y3 (t) = a1 y1 (t) + a2 y2 (t) for any constants a1 and a2

x 1(t)

y1(t)

x 3( t) = a1x 1( t) + a2x 2( t)

x 2 (t )

y3 ( t) = a1 y1( t) + a2 y2 ( t)

Figure 34: A linear system

55

y2 (t)

Example 14: For the system as shown in Figure 35, determine

whether it is a linear system.

x (t)

y ( t ) = x (2t )

Figure 35: A linear system example

56

In this case, the principle of superposition holds, hence it is a

linear system.
x 1 (t)

y 1 (t) = x 1(2t) x 2 ( t)

x 3 (t) = a1x 1( t) + a2 x 2( t)

y 3 (t) = a1 x1 (2 t) + a2 x 2 (2 t)
= a1y 1( t) + a2 y 2( t)

Figure 36: A linear system example

57

y 2 (t) = x 2 (2 t)

Time Invariant

{ A system is time invariant if for any delayed input x(t ; T ),

the output is delayed by the same amount y(t ; T )

x (t )

y ( t)

x1( t ) = x ( t T )

y1( t) = y ( t T )

Figure 37: A time invariant system

58

Example 15: For the system as shown in Figure 38 with

y(t) = x(t) + c, where c is an arbitrary constant, determine
whether it is a time invariant system.

x (t )

y ( t ) = x ( t) + c

Figure 38: A time invariant system example

59

In this case, the system is time invariant.

x ( t)

y ( t) = x ( t ) + c

x1 ( t) = x ( t T )

y1 ( t) = x ( t T ) + c
= y (t T )

Figure 39: A time invariant system example

60

Linear Time Invariant (LTI)

{ A system is linear time invariant if it satises both conditions
of linear and time invariance
{ A LTI system can be analyzed in both time domain and
frequency domain
x 1( t )

y1( t)

x 3( t) = a1x 1 (t T1) + a2 x 2(t T2)

x 2( t)

y 3( t) = a1y1(t T1) + a2 y2(t T2)

Figure 40: A LTI system

61

y 2( t )

Example 16: Determine whether the system given by

y(t) = x(2t) in Example 14 is a LTI system.
From Example 14, the system is linear. However, the system is
not time invariant, hence it is not a LTI system.

x (t)

y ( t) = x (2t)

x1(t) = x ( t T )

y1(t) = x (2t T ) y ( t T )

Figure 41: A non-LTI system example

62

Analysis of DT and CT LTI Systems

Any LTI system can be uniquely dened by its impulse response

(t)

h (t )

[n]

h [n]

Figure 42: Impulse response of a LTI system

63

The output of any LTI system is the convolution of the input

signal and its impulse response

x (t )

h (t)

y (t ) = x (t ) * h (t )

x [n]

h[n]

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

Figure 43: System response of a LTI system

64

The discrete time convolution (convolution sum) is dened as

yn] = xn] hn] =

1
X
m=;1

xm]hn ; m]

The continuous time convolution (convolution integral) is dened

as

y(t) = x(t) h(t) =

65

;1

x( )h(t ;  )d

Example 17: Sketch the waveform of yn] = xn] hn] using the
graphical approach for convolution sum.

x [n]

h [n]

2
1

-1 0

-1 0

Figure 44: Example on convolution sum

66

yn] = xn] hn] =

1
X
m=;1

m
2 h[ ]
1
-2 -1 0

2
1
2 m

-1 0

xm]hn ; m]
h[n m ]

n2 n1

(i)n =-1,

x [m ]

2
1

1
-1 0
(ii)n =0,

2 m

-3 -2 -1

x [m ]

2
1

1
-1 0

2 m

-3 -2

1 m

-1 0

h[ m ]

-1 0
67

1 m

y [n]

h[ 1 m ]

y [n]

-1 0

2 n

(iii)n =1,

x [ m]

2
1

1
-1 0

2m

-2

-1 0

(iv) n =2,

x [ m]

2
1

1
-1 0

h[1 m ]

2m

-2

2 m

-1 0

h[2 m]

-1 0

2 m

1
-1 0

h[3 m ]

1
1

2m

-1 0

3 m

3
2
1

-1 0

(v) n =3,

x [ m]

3
2
1

3
2
1

-1 0

y [ n]

y [ n]

2 n

y [ n]

Figure 45: Solution for example on convolution sum

68

2 n

3 n

Example 18: Show that n ; 2] xn] = xn ; 2] where xn] is shown
in Figure 46.

x [n]
2
1
-2

-1 0

Figure 46: Example on convolution with Delta function

69

Firstly, using the graphical approach. Denote

yn] = n ; 2] xn] =

2
1
-2
(i) n =0,

-1

(ii) n =1,

-1

n 2 n 1 n n+1

3 m

-2

-1

2
1

2 m

. . .

-2

-1 0

x [1 m ]

-1

2 n

y [ n]
2
1

2
1
3 m

y [n ]

2
1

[m 2]

-2

x[n m ]

x[ m]

1
-1

m ; 2]xn ; m]

2
1

[m 2]

m=;1

x [ m]

1
-1

1
X

70

3 m

. . .

-2

-1 0

2 n

(iii)n =2,

x [2 m]

[m 2]

(iv)n=3,

3 m

-1 0

(v) n=4,

3 m

...

-2

-1 0

2 n

3 n

2
1

4 m

..

-1

y [n] = x [n 2 ]
2
1

2
1
3 m

y [n]

x [4 m ]

[m 2 ]

2
1

1
-1 0

3 m

x [3 m ]

[m 2]

1
-1 0

2
1

2
1

1
-1

y [n]

5m

..

-1 0

4 n

Figure 47: Solution for example on convolution with Delta function

71

Alternatively, based on the denition of convolution sum, we have

yn] = n ; 2] xn] =

Since

8
<

m ; 2] =
Hence

yn] =

1
X
m=;1

1
X
m=;1

m ; 2]xn ; m]

1 m = 2
0 m 6= 2:

m ; 2]xn ; m]

= xn ; 2]
72

Example 19: Sketch the waveform of y(t) = x(t) x(t) using the
graphical approach for convolution integral.

x(t)
3
-2

Figure 48: Example on convolution integral

73

y(t) = x(t) x(t) =

3
-2

-2
(i) t<-4,

t -2

;1

x( )x(t ;  )d

x ( )

0
3

-2

x ( )

x(t)
t +2 -2

x ( )

t -2

0
3

x ()

x ( t )

0 t +2

y ( t)

-4
74

(ii) -4< t <0,

x (t)

-2 t +2 0

t -2
(iii) 0<t <4,

x( ) 3

(iv) t >4,

-2

3
0

36

-4

x (t)

0 t -2

-2

x( )

x( )
2 t -2

t +2

0
36

-4

x (t)
-4

t
y (t)

0
36

t +2

y (t)

y (t)

Figure 49: Solution for example on convolution integral

75

Example 20: Sketch the waveform of y(t) = x1 (t) x2 (t) using the
graphical approach for convolution integral.

A1

x 1( t)

x 2 (t )
5

0
A1

Figure 50: Example on convolution integral

76

y(t) = x1 (t) x2 (t) =

A1

;1

x1 ( )x2 (t ;  )d

x 1()

x 2()
5

-5

0
A1

A1

x 1()

x 2( t )
t -5
5

t
0

A1
(i) t <0,

t -5

x1 ()

A1

y (t)

t
0

x 2( t )

A1

77

(ii) 0<t <5,

x 1()

A1

t -5

x 1()
0

10 t

10

5A12

x 2(t )

x 1()

y ( t)
t -5

0
A1

y ( t)
5

0
A1

5A21

A1

t -5

A1

x 2(t )

(iv) t >10,

t
0

(iii) 5<t <10,

A1

y ( t)

5A21

x 2(t )

Figure 51: Solution for example on convolution integral

78

Properties of Convolution
Commutative

x1 n] x2 n] = x2 n] x1 n]

x1 (t) x2 (t) = x2 (t) x1 (t)
Distributive

x1 n] fx2 n] + x3 n]g = x1 n] x2 n] + x1 n] x3 n]

x1 (t) x2 (t) + x3 (t)] = x1 (t) x2 (t) + x1 (t) x3 (t)

79

Associative

x1 n] fx2 n] x3 n]g = fx1 n] x2 n]g x3 n]

x1 (t) x2 (t) x3 (t)] = x1 (t) x2 (t)] x3 (t)
Convolution with Delta function

xn] n ; K0 ] = xn ; K0 ]

x(t) (t ; T0 ) = x(t ; T0 )

80

Example 21: Show that x(t) (t ; T0 ) = x(t ; T0 ) using the

denition of convolution integral.
Based on the denition of convolution integral, we have

y(t) = x(t) (t ; T0 )

Z 1
=
( ; T0 )x(t ;  )d
=

;1
1

;1

( ; T0 )x(t ; T0 )d

= x(t ; T0 )
= x(t ; T0 )
81

;1

( ; T0 )d

Step Response of LTI Systems

The step response is dened as the output of the system with the
unit step function as input signal
Step response of a DT system

sn] = un] hn] =

=

n
X
m=;1

s(t) = u(t) h(t) =

=

;1

m=;1

hm]un ; m]

hm]

Step response of a CT system

Z

1
X

h( )d
82

;1

h( )u(t ;  )d

Example 22: Find the step response of the one-stage RC lter as

shown in Figure 52,; where
the impulse response is given by

1  exp ; t
h(t) = RC
RC u(t).

R
u(t)

s (t)

Figure 52: A simple one-stage RC lter

83

In this case, the step response is given by

s(t) = u(t) h(t)

=

;1

h( )d



 exp ;
u
(

)
d
=
RC
;1 RC
Z t


1

= RC exp ; RC d
0
8

< 1 ; exp ;; t  t  0
RC
= :
0
t < 0:

84

Properties of LTI Systems

Memoryless LTI Systems

{ A LTI system is memoryless if and only if its impulse

response is given by

hn] = cn]
h(t) = c(t)

DT system:
CT system:

where c is an arbitrary constant

{ All memoryless LTI systems simply perform scalar
multiplication on the input

85

Causal LTI Systems

{ A LTI system is causal if and only if its impulse response

satises the following condition

hn] = 0 for n < 0

h(t) = 0 for t < 0

DT system:
CT system:

{ A causal LTI system cannot generate an output before the

input is applied

86

Stable LTI Systems

{ A LTI system is BIBO stable if and only if its impulse response

satises the following condition

1
X

DT system:

jhn]j

<

jh(t)jdt

<

n=;1

CT system:

;1

An example of a stable LTI system

hn] = n un] j j < 1

87

Example 23: Determine whether the system with impulse response

h(t) = exp(;at)u(t) where a > 0 is (i) memoryless, (ii) causal, and
(iii) BIBO stable.
(i) The system is not memoryless since h(t) 6= c(t).
(ii) The system is causal since h(t) = 0 for t < 0.
(iii) The system is BIBO stable since
Z

1
;1

jh(t)jdt

Z
0

exp(;at)dt

= 1=a <

88

System Interconnections
Parallel Connection
h 1 [ n]
x [ n]

h 1 (t)
x (t)

y [ n]

h 1 [ n] + h 2 [ n]

y(t)

h 2 (t) +

h 2 [ n] +

x [ n]

y [ n]

x (t )

h 1 (t) + h 2 (t)

Figure 53: Parallel connection of systems

89

y (t)

Cascade Connection
x [n]

x [n]

h 1[n]

h 2[n]

h 1[n] * h 2[n]

y [n]

x (t)

y [n]

x (t)

h 1( t )

h 1 ( t ) * h 2 ( t)

Figure 54: Cascade connection of systems

90

h 2( t )

y ( t)

y (t)

Example 24: Determine the equivalent impulse response hn] of the

overall system as shown in Figure 55, where h1 n] = un],
h2 n] = un + 2] ; un], h3 n] = n ; 2], and h4 n] =
n un].

h 1 [ n]

x [ n]

h 2 [ n] +

h 3 [ n]
+

y [ n]

h 4 [ n]

Figure 55: Example on interconnections of systems

91

The resultant overall system impulse response is

hn] =

fh1 n] + h2 n]g h3 n] ; h4 n]

= fun] + un + 2] ; un]g n ; 2] ;

n un]
= un + 2] n ; 2] ;
n un]
= un] ;
n un]
= f1 ;
n g un]

92

Dierential and Dierence Equations

Block diagram representation of dierence equation

yn] = xn] ; 3yn ; 1] + 2yn ; 2]

x [ n]

+
+

y [n ]
3

Figure 56: Block diagram representation of dierence equation

93

Block diagram representation of dierential equation

2
5
d
1
d
y(t) = x(t) ; 4  dt y(t) ; 2  dt2 y(t)

x (t )

+
+

y (t )
5
4

d
dt

1
2

d
dt

Figure 57: Block diagram representation of dierential equation

94

Example 25: Find the block diagram representation of dierential

equation for the simple one-stage RC low-pass lter as shown in
Figure 58.

R
x (t)

y (t )

Figure 58: A simple one-stage RC lter

95

In this case, the input and output relation of the lter is given by

d
x(t) = RC  dt y(t) + y(t)
y(t) = x(t) ; RC  dtd y(t)

x (t )

y (t)
RC

d
dt

Figure 59: Block diagram representation of dierential equation for

the RC lter
96

Correlation Function
The correlation function is a mathematical expression of how
correlated two signals are as a function of how much one of them
is shifted
The correlation function between two functions is a function of
the amount of shift
Two types of correlation functions

{ Autocorrelation function
{ Cross correlation function
97

Autocorrelation function
{ The autocorrelation is the correlation of a function with itself
{ For an energy-type signal xn] or x(t)
DT signal:

Rxx m]

CT signal:

Rxx ( )

1
X

n=;1
Z 1

;1

xn]x n + m]

x(t)x (t +  )dt

where x (t) denotes the complex conjugation of x(t).

{ For a power-type signal xn] or x(t)
K
X
1
 n + m]
x

n
]
x
DT signal: Rxx m] = Klim
!1 2K + 1 n=;K
CT signal:

1
Rxx ( ) = lim
T !1 T
98

T=2

;T=2

x(t)x (t +  )dt

For an energy-type signal xn] or x(t)

Ex =
Ex =

DT signal:
CT signal:

Rxx 0]

Rxx (0)

For a power-type signal xn] or x(t)

Px =
Px =

DT signal:
CT signal:

99

Rxx 0]

Rxx (0)

Properties of autocorrelation function

{ The peak of autocorrelation function occurs at the zero shift

DT signal:
CT signal:

Rxx 0]

Rxx (0)




Rxx m]
Rxx ( )

{ Autocorrelation functions are even functions

Rxx m]

DT signal:
CT signal:

=
Rxx ( ) =

Rxx ;m]
Rxx (; )

{ A time shift in the signal does not change its autocorrelation

function, i.e., the autocorrelation functions of x(t) and x(t ; T )

are the same
100

Example 26: Find the autocorrelation function and power of the

sinusoidal signal x(t) = A sin(2 f0 t).
Since x(t) is a power-type signal, the autocorrelation function is
given by
1
Rxx ( ) = lim
T !1 T

T=2

;T=2

x(t)x (t +  )dt

Z T=2
1
2
= Tlim
A
 sin(2 f0 t)  sin(2 f0 (t +  ))dt
!1 T

1
= Tlim
!1 T

;T=2
T=2

;T=2

A2 cos(2 f  ) ; cos(2 f (2t +  ))] dt

0
0
2

2
A
= 2  cos(2 f0  )

101

The power of signal x(t) is given by

Px = Rxx (0)
2
A
= 2

Alternatively, based on the denition of power, we have

Z T=2
2
P = lim 1
jx(t)j dt
x

T !1 T ;T=2
1 Z T0 =2

2
jx(t)j dt
= T
0 ;T0 =2
Z T0 =2
= T1
jA sin(2 f0 t)j2 dt
0
2

A
= 2

;T0 =2

102

Example 27: Find the autocorrelation function and power of the

sinusoidal signal y(t) = A sin 2 f0 (t ; T )], where T is an
arbitrary constant delay.
Denote  = 2 f0 T , we have

y(t) = A sin(2 f0 t ; 2 f0 T )
= A sin(2 f0 t ; )
Since y(t) is a power-type signal, the autocorrelation function is
given by
1
Ryy ( ) = lim
T !1 T
103

T=2

;T=2

y(t)y (t +  )dt

1
= Tlim
!1 T

T=2

;T=2

A2  sin(2 f0 t ; )  sin(2 f0 (t +  ) ; )dt

Z T=2
2
A
1
cos(2 f0  ) ; cos(2 f0 (2t +  ) ; 2)] dt
= Tlim
!1 T ;T=2 2
2
A
= 2  cos(2 f0  )

Comparing with the results with Example 26, we conclude that

the autocorrelation functions of x(t) and x(t ; T ) are the same.

104

Example 28: Find the autocorrelation function of the signal xn]

as shown in Figure 60 using the graphical approach.

x [n]
2
1
0

Figure 60: Example on autocorrelation function of a DT signal

105

Since xn] is an energy-type signal,

Rxx m]

2
1
-1 0

1
X

n=;1

xn]x n + m] =

x [ n]
1

2 n

2
1
0

m 1 m m+1

1
X
n=;1
x [ n+m ]
1

(i)m =2,
2
1
-1

x [ n]
1

(ii) m=1,
x [ n]
2
1
-1 0
1

2
1
2 n

2 n

-2

-2

-1

2
1
-1 0

x [ n +2]
1

2 n

-2

106

2 n
6
4
2
-1 0

6
4
2

x[ n +1]
2 n

xn]xn + m]

-2

-1 0

R xx [ m]

.m
2

R xx [ m]

.m
2

(iii) m=0,
2
1
-1 0

x [n]
1

2 n

-2

2
1
-1 0

(iv) m =-1,
2
1

x [n]

-1 0

-2

-1 0

(v) m=-2,
2
1
-1 0

x [n]
1

2
1
2 n

-1 0

2 n

4
2
-2 -1 0

2 m

2 n

-2

3 n

6 R xx [ m]
4
2
-2 -1 0
1 2 m

x [n 2]
1

6 R xx [ m]
4
2

x [n 1]

2
1
2 n

[ m]
R
xx
6

x [n]

-1 0

2 m

Figure 61: Solution for example on autocorrelation function

107

Cross correlation function

{ The cross correlation is the correlation of two dierent functions
{ For energy-type signals xn] and yn] (or x(t) and y(t))
DT signal:

Rxy m]

CT signal:

Rxy ( )

1
X

n=;1
Z 1

;1

xn]y n + m]

x(t)y (t +  )dt

{ For power-type signals xn] and yn] (or x(t) and y(t))
DT signal:
CT signal:

K
X
1
 n + m]
Rxy m] = lim
x

n
]
y
K !1 2K + 1 n=;K
Z T=2
1
 (t +  )dt
x
(
t
)
y
Rxy ( ) = lim
T !1 T

;T=2

108

Example 29: Find the cross correlation function between the two
signals x(t) = exp(j 2 f0 t) and y(t) = exp(j 2 2f0 t).
Since x(t) and y(t) are power-type signals,
Z T=2
1
 (t +  )dt
x
(
t
)
y
Rxy ( ) = lim
T !1 T

1
= Tlim
!1 T

;T=2
T=2

;T=2

exp(j 2 f0 t)  exp;j 2 2f0 (t +  )]dt

Z T=2
1
= Tlim
exp(;j 2 f0 t)  exp(;j 4 f0  )dt
!1 T

;T=2

1
= exp(;j 4 f0  ) Tlim
!1 T

T=2

= 0

109

;T=2

cos(2 f0 t) ; j sin(2 f0 t)] dt

Example 30: Find the cross correlation function between the two
signals x(t) and y(t) as shown in Figure 62 using the graphical
approach.
x (t)

y (t )

2
2

-2

-2

0
-2

Figure 62: Example on cross correlation function of CT signals

110

Since x(t) and y(t) are energy-type signals,

Rxy ( )

=
2

-2
(i) >4,

;1

x(t)y (t +  )dt =

;1

x(t)y(t +  )dt
y (t + )
2

+2

-2

x(t)

0
-2

x ( t)

0
-2

y (t + )

-2

y (t + )

x(t)

(ii) 2< <4,

+2

0
-2

R xy()
t

R xy ()

+2

0
-8

111

(iii) 0<<2,

y (t +)

R xy ()

x(t)

+2

2 -2

x(t)

2 0

-2

x(t)
-2

(vi) <-4,

-2

2
0
-2

0
-2

R xy ()

+2

-2
(v) -4< <-2,

-8

-2
(iv) -2< <0,

t
y (t +)

-2

-8

y (t +)
2 2

+2

-4

-2

R xy ()

-8

()
8 R xy

y (t +)

x(t)
2

+2
t

-4

-2

0
-8

Figure 63: Solution for example on cross correlation function

112

NANYANG TECHNOLOGICAL UNIVERSITY

SCHOOL OF ELECTRICAL & ELECTRONIC ENGINEERING
EE2010/IM2004 SIGNALS AND SYSTEMS
TUTORIAL 1
Q1.1: Determine the even and odd components of the DT function x[n ] = n + ( 1) n . Sketch the
waveforms of xe[n], xo[n], and x[n].

Q1.2: Determine whether the following signals are energy-type or power-type signals:
(a)

e|t| , for 2 t 2,
x (t ) =
0, otherwise.

(b)

y[n ] = ( 1) n .

Q1.3: A sinusoidal signal x (t ) = 2 sin(20t ) is passed through a half-wave rectifier circuit to produce:

x (t ), if x (t ) > 0,
y (t ) =
0, otherwise.
(a) Sketch the waveforms of x(t) and y(t), respectively.
(b) Determine the energy and power levels of x(t) and y(t), respectively.

Answers :
Q1.1 : xe [n ] = ( 1)n , xo [n ] = n.
Q1.2 : (a) E x = e 4 1, Px = 0. (b) E y = , Py = 1.
Q1.3 : (b) E x = , Px = 2, E y = , Py = 1.
1

NANYANG TECHNOLOGICAL UNIVERSITY

SCHOOL OF ELECTRICAL & ELECTRONIC ENGINEERING
EE2010/IM2004 SIGNALS AND SYSTEMS
TUTORIAL 2
Q2.1: Assume that v (t ) = sin(2f 0 t ), sketch the following waveforms and evaluate

y (t )dt.

t
, where T0 = 1 / f 0 .
w(t ) = v (t ) rect
2T0

(a)

(b)

x (t ) = sgn(w(t ) ).

(c)

y (t ) =

x(t ) (t nT

n =

Q2.2: Sketch the waveforms of x (t ) =

/ 4).

n = 1

and T0 = 2.

t +4
, where v (t ) = t[u(t ) u(t T0 )]
2

v(t nT ) and y (t ) = x
0

Q2.3: Assuming that the signal v(t) is an energy-type signal and its energy is denoted as Ev, determine
the energy levels of the following signals as a function of Ev.
(a)

x (t ) = 3v (t ).

(b)

y (t ) = v (t 3).

Answers :
Q 2.1 :

y (t )dt = 0.

Q 2.3 : (a) E x = 9 E v . (b) E y = E v .

2

NANYANG TECHNOLOGICAL UNIVERSITY

SCHOOL OF ELECTRICAL & ELECTRONIC ENGINEERING
EE2010/IM2004 SIGNALS AND SYSTEMS
TUTORIAL 3

1, for n = 1,0,1.
0, otherwise.

Q3.1: Evaluate the convolution sum y[n ] = x[n ] x[n ] where x[n ] =

Q3.2: For the system as shown in Figure Q3.2, evaluate the system output y(t) where
t 1
t 3
t 2
x1 (t ) = A rect
A rect
and x 2 (t ) = A rect
.

2
2
4

Figure Q3.2.

Q3.3: Determine the properties of the system shown in Figure Q3.3 in terms of linearity and time
invariance.

Figure Q3.3.

Answers :
Q 3.3 : The system is linear but not time invariant.
3

NANYANG TECHNOLOGICAL UNIVERSITY

SCHOOL OF ELECTRICAL & ELECTRONIC ENGINEERING
EE2010/IM2004 SIGNALS AND SYSTEMS
TUTORIAL 4
Q4.1: Radio signals can travel through a wireless channel by more than one path, with different time
delays and attenuations (known as channel fading). Consider a three-path case with a system impulse
response given by
k

1
h(t ) = (t 3k ).
k =0 2
t -1
Assume that the input signal is given by x (t ) = t rect
.
2
2

(a) Determine whether the system is memoryless, causal, and stable.

(b) Determine the system output y(t) and sketch its waveform.
(c) Express the energy of y(t) as a function of the energy of x(t).
Q4.2: Evaluate the step response of the system with impulse response given by h (t ) = 2e 2|t| .
Q4.3: (a) Find the equivalent impulse response h[n] of the overall system as shown in Figure Q4.3,
where h1 [n ] = [n 4], h2 [n ] = [n 2], and h3 [n ] = [n + 1].
(b) Determine whether the overall system is memoryless, causal, and stable.
(c) Determine the step response of the overall system.

Figure Q4.3.

Answers :
Q 4.1 : (a) The system is not memoryless, it is causal and stable. (b) E y =

21
Ex .
16

2 e 2 t , for t 0,
Q 4.2 : s(t ) = 2 t
e , for t < 0.
Q 4.3 : (a) h[n ] = [n 1] + [n 3]. (b) The system is not memoryless, it is causal and stable.
(c) s[n ] = u[n 1] + u[n 3].
4

NANYANG TECHNOLOGICAL UNIVERSITY

SCHOOL OF ELECTRICAL & ELECTRONIC ENGINEERING
EE2010/IM2004 SIGNALS AND SYSTEMS
TUTORIAL 5
Q5.1: Show that the cross-correlation function between any energy-type signal x[n] and the delta
function [n ] is equal to x[-m].

n
Q5.2: Find the cross-correlation function between the two signals x[n ] = rect and y[n ] = [n 1].
2

Figure Q5.2.

Q5.3: Consider two complex-valued signals given by x (t ) = 3e j 20t and y (t ) = 4e j 40t .

(a)
(b)
(c)
(d)

Sketch the amplitude plots of x(t) and y(t), respectively.

Determine the power levels of x(t) and y(t), respectively.
Find the cross-correlation function of x(t) and y(t).
Comment on the result obtained in part (c).

Answers :
Q5.3 : (b) Px = 9, Py = 16. (c) R xy ( ) = 0.
5