You are on page 1of 18

离散傅里叶变换

( Discrete
( Discrete Fourier
Fourier Transform,
Transform, DFT
DFT )

离散时间线性系统 数字信号处理
Outline

Sampling in
time domain
Introduction

Time Domain
Analysis of
Discrete Time
Signal and Frequency
System Domain and
z Domain

离散时间线性系统 2 数字信号处理
Outline
Sampling in
Time Domain
Discrete Fourier Series
Introduction 绪论

Time domain ( DFS )


analysis of discrete
time signals

Time domain Discrete time Fourier


analysis of discrete transform ( DTFT )
time Systems
Discrete Fourier
Frequency domain
analysis of
transform ( DFT )
discrete time
Systems Fast Fourier transform
( FFT )
离散时间线性系统 3 数字信号处理
Method for finding the spectrum according to th
e time domain signal
Periodic signal Aperiodic signal
x (t ) 
FS
 X ( jk 0 ) x (t ) 
FT
 X ( j )
Continuou X ( j)
nx

s time x t X ( jk0 ) x(t )


signal
t k 
t

x ( n) 
DFS
 X (k ) x(n) 
DTFT
 X (e j )
X (k )
Discrete x(n) x(n) X (e j )
time signal
n k n 

process discrete data


Features of digital
processing equipment
Limited storage space

离散时间线性系统 4 数字信号处理
Time Domain : x(n) x (n)
Relationship

Sequence with finite length of N Periodic sequence with period
N
x(n)

0 n
~
x ( n)

… …
0 n
x(n)  x (n) RN (n) Main value sequence of x ( n)

x (n)   x(n  rN )  x((n))
r 
N
period extension of x(n)

离散时间线性系统 5 数字信号处理
Frequency X (k ) X (k )
Domain : Relationship ?

Spectrum with finite length of N Periodic spectrum with period N

X (k )  X (k ) RN (k )

X (k )   X (k  rN )  X ((k ))
r 
N

 1 n  0,1,..., N  1
RN (n)  
0 其他
remainder operation , namely n to N to
  n   The
N find the remainder
x   8   8 
8 to 8 remainder is equal to 0
 x   8   8  x(0)
x   9   8  x   8  1  8 
9 to 8 remainder is equal to 1
 x   9   8  x(1)

离散时间线性系统 6 数字信号处理
x(n) x (n)  x (n) RN ( n)
X (k )  X (k ) RN (k )
0~ n
x ( n)

… …
x (n)  x(( n)) N
0 n X (k )  X ((k )) N

N 1 2
j nk
X (k )   x (n)e N
DFS n0
2
Finite-length sequence 1 N 1  j nk
frequency domain x ( n)   X ( k )e N
analysis method N k 0

x (n)  x(( n)) N X (k )  X (k ) RN ( k )


DFS
x(n) x (n) X ( k ) X (k )

离散时间线性系统 7 数字信号处理
7
DFT positive and inverse transformation of finit
e-length sequences
N 1
X (k )  DFT [ x(n)]   x (n)WNnk 0  k  N 1
n 0

1 N 1
x( n)  IDFT [ X (k )]   X (k )WN nk 0  n  N  1
N k 0
N 1
or: X (k )   x (n)WNnk RN (k )  X (k ) RN (k )
n 0

1 N 1 
x(n)   X (k )WN nk RN (n)  x (n) RN (n)
N k 0
2
j
where: WN  e N

离散时间线性系统 8 数字信号处理
DFT
N 1
x (n) X (k )  DFT [ x (n)]   x( n)WNnk X (k )
n 0 Discrete spectrum
Sequence with
finite length of N with finite length of
N
N discrete values N discrete values
(time domain DFT (discrete
sequence) spectrum)

x ( n) X (k )

0 1 2 3 4 N-1 01 2 3 4 N-1 k
n
离散时间线性系统 99 数字信号处理
Relationship between N-point DFT and DTFT
 N 1
X (e j )  DTFT  x(n)    x ( n ) e  j n
  x ( n ) e  j n

n  n 0
N 1 N 1 2
j nk
X (k )  DFT [ x(n)]   x( n)W nk
N   x( n)e N

n 0 n0

 X (e j ) 2 k 0k  N
 
X (k )   N


0 otherwise

The N-point DFT of the finite-length aperiodic


sequence x(n) is an equally spaced sampling of N
points of the D(n) DTFT over the interval [0, 2π].
离散时间线性系统 10 数字信号处理
Relationship between N-point DFT and DTFT

X ( ) 2
2 2
3 1 2
0  
N 8
N
  4
 k=0
3   3
2
5 7(k=N-1)
6
 X ( e ) 2 k
j
0k  N
 
X (k )   N

 0 otherwise

2 k 2 k
  k  0,1,..., 7
N 8
离散时间线性系统 11 数字信号处理
Example
Problem: Find the 8-point and 16-point DFT for the given
sequence x(n)  R4 (n)
Solution:find the DTFT for x  n 
 3
Xe j
   x  n e  j n
 e  j n

1  e  j 4
n  n 0 1  e  j

e  j 2  e j 2  e  j 2 
 
j  j 2 j


e 2
 e  e 2

 
3
j  sin  2 
e 2
sin   / 2 

离散时间线性系统 12 数字信号处理
Example—find 8-point DFT (Method 1)
find 8-point DFT for x  n  N 8

X  k   X  e j  2
 k
8

 2 
3 sin 2 k
j  k
e 24  8  The N-point DFT of x(n) is an
equally spaced sampling of
 1 2  N points of the DTFT of x(n)
sin   k over the interval [0, 2π]
2 8 
 
3 sin  k
 j k
e 8 2 
 
sin  k 
8 
离散时间线性系统 13 数字信号处理
Example—find 8-point DFT (Method 2)
find 8-point DFT for x  n  N 8
7 3 2
j kn
X  k    x(n)W 8
nk
 e 8

n 0 n 0
  
j k j k j k
 j k
1 e e 2
e 2
e 2
 
 
  
j k j k j k j k
1 e 4
e e 8 8
e 8

 
3 sin  k
 j k
e 8 2 
 
sin  k 
8 
k  0,1, , 7

离散时间线性系统 14 数字信号处理
Example—find 16-point DFT (Method 1)
find 16-point DFT for x  n  N  16

X  k   X  e j  2
 k
16

 2 
3 2 sin  2  k
j  k
 16 
 e 2 16
 1 2 
sin   k
 2 16 
 
3 sin  k 
e
 j k
16 4 
 
sin  k 
 16 
离散时间线性系统 15 数字信号处理
Example—find 16-point DFT (Method 2)
find 16-point DFT for x  n  N  16
15 3 2
j kn
X  k    x (n)W 8
nk
 e 16

n 0 n 0
   
j k j k j k j k
1 e 2
e 4
e 4
e 4
      
j k j k j k j k
1 e 8
e 16
e 16
e 16

 
3 sin  k
 j k
 e 16 4 
 
sin  k 
 16 
k  0,1, ,15
离散时间线性系统 16 数字信号处理
References
1 、程佩青,《数字信号处理教程(第四版)》,清华大学出版社, 2015.
2 、唐向宏、孙闽红,《数字信号处理——原理、实现与仿真(第二版)》,高
等教育出版社, 2014.
3 、唐向宏、岳恒立、孙闽红,《数字信号处理实践教程》,高等教育出版社,
2013.
4 、 A.V.Oppenheim 、 R.W. Schafer ,《离散时间信号处理(第 3 版)》(英
文版) , 电子工业出版社, 2011.
5 、 J. G.Proakis , D.G. Manolakis ,《数字信号处理:原理、算法与应用(第
4 版)》(英文版),电子工业出版社, 2013.
6 、 V.K. Ingle 、 J.G. Proakis ,《数字信号处理——应用 MATLAB (第三
版)》(英文影印版),科学出版社, 2016.
7 、 S. Poornachandra 、 B. Sasikala ,《数字信号处理》(英文影印版),科
学出版社, 2012.
8 、 James McNames 教授的讲义 http://web.cecs.pdx.edu/~mcnames/

离散时间线性系统 17 数字信号处理
南京信息工程大学

Thank you

离散时间线性系统 18 数字信号处理

You might also like