Scribd
Upload
99 views

Digital Signal Analysis and Applications

The document discusses various techniques for analyzing aperiodic signals, including the Fourier transform. The Fourier transform represents an aperiodic signal as a continuous sum of sinusoids, unlike the Fourier series which requires the signal to be periodic. The Fourier transform of a signal is its frequency spectrum, obtained by taking the integral of the signal multiplied by complex exponentials. Convolution in the time domain results in multiplication in the frequency domain. The discrete Fourier transform (DFT) represents a discrete-time signal as a sum of sinusoids, with frequencies that are integer multiples of the fundamental frequency.

Uploaded by

GauravMishra
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
99 views

Digital Signal Analysis and Applications

The document discusses various techniques for analyzing aperiodic signals, including the Fourier transform. The Fourier transform represents an aperiodic signal as a continuous sum of sinusoids, unlike the Fourier series which requires the signal to be periodic. The Fourier transform of a signal is its frequency spectrum, obtained by taking the integral of the signal multiplied by complex exponentials. Convolution in the time domain results in multiplication in the frequency domain. The discrete Fourier transform (DFT) represents a discrete-time signal as a sum of sinusoids, with frequencies that are integer multiples of the fundamental frequency.

Uploaded by

GauravMishra
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 13

Digital Signal Analysis and Applications

Kishore Prahallad Assistant Professor, IIIT Hyderabad. Email: [email protected] http://www.iiit.ac.in/kishore

Contents
Contents 1 Lecture: Dealing with aperiodic signals 1.1 Fourier transform . . . . . . . . . . . . . . . 1.1.1 From 0 to . . . . . . . . . . . . . . 1.1.2 Fourier transform (FT) pair . . . . . 1.1.3 Convolution and FT . . . . . . . . . Example problems . . . . . . . . . . 1.1.4 FT of a sequence of impulses . . . . FT of a periodic signal . . . . . . . . Application to sequence of impulses 1.2 Graphical development . . . . . . . . . . . . 1.3 Discrete-time Fourier transform (DTFT) . . 1.4 Discrete Fourier transform (DFT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 3 . 4 . 5 . 6 . 6 . 7 . 7 . 7 . 8 . 10 . 11 . 12

Chapter 1 Lecture: Dealing with aperiodic signals


So far, we discussed approximation of x(t) using cosine and sine signals. We derived the approximation in terms of exponential series as shown below.

x( t ) = 1 ck = T

ck ejk0 t
k= T

(1.1) x(t)ejk0 t dt.

The equations for discrete Fourier series are as follows


K 1

x [ n] =
k =0

ck ejk2 N 1 N
N 1

(1.2) x[n]ejk2 N ,
n

ck =

n=0

where N is the period of the signal. An important point to note in the above equations is that the Fourier series require the knowledge of 0 (or T ). This also implies that the signal x(t) should be periodic. The question is how to approximate aperiodic signals? 3

CHAPTER 1. LECTURE: DEALING WITH APERIODIC SIGNALS

1.1

Fourier transform
x(t)i s N on- peri odi c

Amplitude

x(t)

0 x(t) i s P eri odi c

Amplitude

x(t)

0 time

2T

Figure 1.1: (a) aperiodic signal x(t) and (b) periodic signal x (t) by repeating the aperiodic signal Let x(t) be an aperiodic signal. Using x(t), one could create a (pseudo) periodic signal x (t) as shown in Fig. 1.1. Following Fourier series,

x (t) = 1 ck = T

ck ejk0 t dt
k= T

(1.3) x (t)ejk0 t dt.

CHAPTER 1. LECTURE: DEALING WITH APERIODIC SIGNALS

1.1.1 From 0 to
Follow 1 ck = T
T

x (t)ejk0 t dt.
0

Note that x(t) = x (t) 0 < t < T, else x(t) = 0. Using this equality ck = 1 T 1 ck = T
T

x(t)ejk0 t dt
0

(1.4) x( t ) e
jk0 t

dt

Let

X ( ) =

x(t)ejt dt (1.5)

1 ck = X (k0 ) T

Following Eq. (1.5), Fourier Coefcients are samples of X ( ) taken at multiples of 0 . Using Eq. (1.5) in Eq. (1.3), 1 x (t) = T

X (k0 )ejk0 t
k=

1 = 2

(1.6) X (k0 )ejk0 t 0

k=

As T ; x(t) x (t); 0 0;

As 0 0, it implies that Eq. (1.6) converges to an integral, following Riemann summation. x( t ) = 1 2

X ( )ejt d

(1.7)

CHAPTER 1. LECTURE: DEALING WITH APERIODIC SIGNALS

1.1.2 Fourier transform (FT) pair


x( t ) = X ( ) =

1 2

X ( )ejt d

(1.8)
jt

x( t ) e

dt

1.1.3 Convolution and FT


Let denotes the convolution operator. The convolution of x(t) and h(t) y ( t ) = x( t ) h ( t )

h( )x(t ) d

x( t ) =

1 X ( )ej(t ) d 2 1 = X ( )ejt ej d 2 1 y (t) = h( ) X ( )ejt ej d d 2 1 h( )ej d X ( )ejt d y (t) = 2 1 h( )ej d X ( )ejt d y (t) = 2 1 H ( )X ( )ejt d = 2 Y ( ) = H ( )X ( )

(1.9)

Convolution in time domain leads to multiplication in frequency domain and vice-versa, i.e., convolution in frequency ( ) domain leads to multiplication in time domain.

CHAPTER 1. LECTURE: DEALING WITH APERIODIC SIGNALS Example problems

1.1.4 FT of a sequence of impulses


Let us say, we want to derive the Fourier transform for a sequence of impulses with a time period T , i.e.,

x( t ) =
k=

(t kT ).

It could be easily realised that sequence of impulses is a periodic signal and Fourier series coefcients can be generated. In order to derive the Fourier transform, one could know the FT of a periodic signal as well. So far, we dealt with Fourier transform of an aperiodic signal. However, it would be helpful to have Fourier transform for a periodic signal as well as, to have a unied framework. FT of a periodic signal Let X ( ) = 2 ( 0 ) (1.10)

The corresponding x(t) can be obtained by the Fourier transform pair.


1 x( t ) = X ( )ejt d 2 1 = 2 ( 0 )ejt d 2 = ej0 t

(1.11)

CHAPTER 1. LECTURE: DEALING WITH APERIODIC SIGNALS 8 If denotes the Fourier transform of an LTI system then
x(t) X ( ) ej0 t 2 ( 0 ) ejk0 t 2 ( k0 ) ck ejk0 t 2ck ( k0 )

(1.12)

ck e
k=

jk0 t

2
k=

ck ( k0 ) ck ( k0 )
k=

X ( ) = 2

Application to sequence of impulses

x( t ) = 1 T 1 = T 1 = T 2 X ( ) = T ck =

(t kT )
k= T

x(t)ejk0 t dt
0 T

(t)ejk0 t dt
0

(1.13)

( k
k=

2 ) T

CHAPTER 1. LECTURE: DEALING WITH APERIODIC SIGNALS

..

..

3T

2T

2T

3T

..

..

2/T

2/T

Figure 1.2: (a) Impulse train x(t) and (b) Fourier transform of impulse train

CHAPTER 1. LECTURE: DEALING WITH APERIODIC SIGNALS

10

1.2

Graphical development

Below is the graphical development of Fourier transform (FT), Discrete-time Fourier transform (DTFT) and Discrete Fourier transform (DFT).

Figure 1.3: (a) Graphical development of DFT

CHAPTER 1. LECTURE: DEALING WITH APERIODIC SIGNALS

11

1.3

Discrete-time Fourier transform (DTFT)


x( t ) = X ( ) =

1 2

X ( )ejt d

(1.14)
jt

x( t ) e

dt

Let us say we have x[n] instead of x(t).

X ( ) =
n=

x[n]ejn x[n]ej (+2)n


n=

X ( + 2 ) = =
n=

(1.15)

x[n]ejn ej 2n = X ( )

X ( ) is periodic with period 2 1 x [ n] = 2


2

X ( )ejn d
0

(1.16)

CHAPTER 1. LECTURE: DEALING WITH APERIODIC SIGNALS Continuous Discrete Periodic FS DFS

12

x( t ) = ck = Aperiodic x( t ) =
1 2 1 T k= T

ck ejk0 t x(t)ejk0 t dt

K 1

x [ n] = ck =
1 N k=0 N 1 n=0

ck ejk2 N
n

x[n]ejk2 N

FT

DTFT x [ n] =
1 2 2

X ( )ejt d

X ( )ejn d
0

X ( ) =

x( t ) e

jt

dt

X ( ) =
n= N 1

x(n)ejn DFT

x [ n] =

1 N

X [k ]ej N kn
2

k=0 N 1 n=0

X [k ] =

x[n]ej N kn

1.4

Discrete Fourier transform (DFT)


Let N = 2 2 = N = k 1 x [ n] = 2 1 = 2 = x [ n] = 1 N 1 N
2

X ( )ejn d
0 N 1

(1.17) X [k ]e
jkn

k=0 N 1

X [k ]ej N kn
k=0 N 1

X [k ]ej N kn ,
k=0

CHAPTER 1. LECTURE: DEALING WITH APERIODIC SIGNALS 13 where X [k ] is written as X [k ]. Also note that, as is discretized , x[n] will become periodic.
N 1

X [k ] =
n=0 N 1

x[n]ejkn (1.18) x [ n] e
n=0
j 2 kn N

You might also like