Vous êtes sur la page 1sur 8

# COMMUNICATION SYSTEMS

Lecture # 4
10th Feb 2007

Instructor
WASEEM KHAN

## Continuous-time, Discrete-time and

Digital Signals
6

4 Continuous-time
x(t)

0
-1 0 1 2 3 4 5 6
6
x(nTs)

0
Discrete-time
0 Ts 2Ts 3Ts 4Ts 5Ts 6Ts

4.5
xQ(nTs)

3
1.5
Quantized
0
0 Ts 2Ts 3Ts 4Ts 5Ts 6Ts
Time (seconds)

## A quantized signal, when represented by a sequence of bits is called digital.

1
More Examples of Discrete-time Signals

Discrete-time Signals
CT x(t)
DT x(nTs)
( t is replaced by nTs where n = .., -2, -1, 0, 1, 2, 3, ..)
DT signals are usually represented as x(n) or x[n].
Inverse of sampling time period is called sampling frequency given
by
fs = 1/ Ts
According to Nyquist criteria, sampling frequency should be at least
twice of the highest frequency content in the signal to be sampled.
Otherwise some frequency contents will be lost and some will be
corrupted. This phenomenon is called aliasing.
In discrete-time the highest signal frequency is given by fs / 2 Hz.

2
Sampling of Analog Signals and Aliasing
1

0.5
fsam = 50 Hz
0
fsig = 10 Hz
-0.5

-1
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4
Time (milliseconds)

0.5
fsam = 12.5 Hz
0

-0.5 fsig = 10 Hz
-1
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4

Time (milliseconds)

Aliasing

## Sampling of analog signals behaves as filtering operation; it

removes all frequency components above fs/2.
Under-sampling also causes aliasing; high frequency components
are mapped onto low frequency components.
Aliasing causes not only loss of high frequency components but also
corruption of low frequency components.
If the original signal is to be reconstructed from its sampled version,
even the low frequency components can t be, in the same form as
they were before sampling.

3
Discrete-time Fourier Transform
DTFT is applied to discrete-time signals

j t
CTFT X x (t ) e dt

j n
DTFT X x [ n ]e
n

1
Inverse DTFT x[ n ] X ( )e j n d
2

## Discrete-time Fourier Transform

DTFT is periodic over 2 .
Proof :
j n
X( ) x[n]e
n

j( 2 )n
X( 2 ) x[n]e
n

j n j2 n
x[n]e e
n

j n
x[n]e [cos(2 n) j sin(2 n)]
n

j n
x[n]e [1 0]
n

j n
x[n]e X( )
n

4
Effect of Time Domain Sampling in
Frequency Domain
We have proved that DTFT is periodic over 2 .
When a continuous-time signal is sampled with a frequency fs, its
continuous-time Fourier transform is copied onto the frequencies ± fs, ±
2fs, ± 3fs ,
=2 corresponds to fs.
sampling
x (t ) x[n]
|X( )| |X( )|

-fa 0 fa -fs 0 fs

-2 2

## Effect of Aliasing in Frequency Domain

X( )
Over-sampling

-fs 0 fs

X( )
Under-sampling

-fs - fs /2 0 fs /2 fs

5
Some Properties of DTFT
Linearity:
If f [n] F ( ) and g [n] G( )
then af [n] + bg [n] aF ( ) + bG ( )

Time shifting:
If f [n] F( )
then f [n-n0] exp(-j n0) F ( )

If f [n] F( )

jdX
then nx [ n ]
d

## Frequency shifting (modulation)

If f [n]
F( )
then exp[ j 0 n] f [n] F( - 0)

Time reversal
If f [n] F( )
then f [- n] F (- )

6
Important DTFT Pairs
x[n] X( ) Remarks
[n] 1
1 2 [ ]
cos [ 0n] [ ( - 0) + ( + 0)
sin [ | | ,
0n] -j [ ( - 0) - ( + 0)
u[n] 1 | 0|
j
1 e
exp (j 0n) 2 ( - 0)

## an u[n], |a| < 1 1

j
1 ae

DTFT Problems
Find DTFT of
1 M /2 n M /2
x[n]
0 elsewhere

7
This document was created with Win2PDF available at http://www.daneprairie.com.
The unregistered version of Win2PDF is for evaluation or non-commercial use only.