Scribd
Upload
11 views19 pages

Discrete Fourier Transform

The document provides an overview of the Discrete Fourier Transform (DFT) and its fast algorithms, specifically the Fast Fourier Transform (FFT). It includes analysis and synthesis equations, computational complexities, and discusses techniques like decimation-in-time and bit-reverse reordering. The author, Liang Dong from Baylor University, presents these concepts in the context of digital signal processing.

Uploaded by

Mohamed shabana
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
11 views19 pages

Discrete Fourier Transform

The document provides an overview of the Discrete Fourier Transform (DFT) and its fast algorithms, specifically the Fast Fourier Transform (FFT). It includes analysis and synthesis equations, computational complexities, and discusses techniques like decimation-in-time and bit-reverse reordering. The author, Liang Dong from Baylor University, presents these concepts in the context of digital signal processing.

Uploaded by

Mohamed shabana
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 19

ELC 4351: Digital Signal Processing

Liang Dong

Electrical and Computer Engineering


Baylor University
liang [email protected]

April 6, 2017

Liang Dong (Baylor University) Discrete Fourier Transform April 6, 2017 1 / 19


Discrete Fourier Transform

1 Fourier Transform and Discrete Fourier Transform

2 Fast Fourier Transform Algorithms

Liang Dong (Baylor University) Discrete Fourier Transform April 6, 2017 2 / 19


The Fourier Transform of Discrete-Time Aperiodic Signals

Analysis Equation
∞
X
X (ω) = x(n)e −jωn , ω ∈ [−π, π) or ω ∈ [0, 2π)
n=−∞

Synthesis Equation
Z π
1
x(n) = X (ω)e jωn dω
2π −π

X (ω) is periodic with period 2π.

Liang Dong (Baylor University) Discrete Fourier Transform April 6, 2017 3 / 19


The Discrete Fourier Transform (DFT)

N-point DFT.
Analysis Equation
N−1
X k
X (k) = x(n)e −j2π N n , k = 0, 1, 2, . . . , N − 1
n=0

Synthesis Equation
N−1
1 X k
x(n) = X (k)e j2π N n , n = 0, 1, 2, . . . , N − 1
N
k=0

N samples of the Fourier transform at N equally spaced frequencies.


ωk = 2πk
N , k = 0, 1, 2, . . . , N − 1.

Liang Dong (Baylor University) Discrete Fourier Transform April 6, 2017 4 / 19


A Fast Algorithm for the DFT

Analysis Equation
N−1
X
X (k) = x(n)WNkn , k = 0, 1, 2, . . . , N − 1
n=0

Synthesis Equation
N−1
1 X
x(n) = X (k)WN−kn , n = 0, 1, 2, . . . , N − 1
N
k=0

where, WN = e −j2π/N .

Liang Dong (Baylor University) Discrete Fourier Transform April 6, 2017 5 / 19


A Fast Algorithm for the DFT

Analysis Equation
N−1
X
X (k) = x(n)WNkn , k = 0, 1, 2, . . . , N − 1
n=0

To calculate one frequency sample (each k) in the analysis equation


(direct Fourier transform), we need N complex multiplications and N − 1
complex additions.

For all N frequency samples, we need a total of N 2 complex


multiplications and N(N − 1) complex additions.

Liang Dong (Baylor University) Discrete Fourier Transform April 6, 2017 6 / 19


A Fast Algorithm for the DFT

Analysis Equation
N−1
X
X (k) = x(n)WNkn , k = 0, 1, 2, . . . , N − 1
n=0

For all N frequency samples, we need a total of 4N 2 real multiplications


and N(4N − 2) real additions.

Liang Dong (Baylor University) Discrete Fourier Transform April 6, 2017 7 / 19


A Fast Algorithm for the DFT

Analysis Equation
N−1
X
X (k) = x(n)WNkn , k = 0, 1, 2, . . . , N − 1
n=0

The computation complexity of the DFT is proportional to N 2 .

As a comparison, the computational complexity of the FFT is proportional


to N log N.

Liang Dong (Baylor University) Discrete Fourier Transform April 6, 2017 8 / 19


Periodicity of WNkn

Use the periodicity of the sequence WNkn to reduce computation.


2π
WNkN = e −j N
kN
= e −j2πk = 1

k(N−n)
WN = WN−kn = (WNkn )∗
(complex conjugate symmetry)

k(n+N) (k+N)n
WNkn = WN = WN
(periodicity)

Liang Dong (Baylor University) Discrete Fourier Transform April 6, 2017 9 / 19


Decimation-in-Time Fast Fourier Transform (FFT)

Considering N an integer power of 2, i.e. N = 2ν .

N−1
X
X (k) = x(n)WNnk , k = 0, 1, . . . , N − 1
n=0
X X
= x(n)WNnk + x(n)WNnk
n even n odd
N/2−1 N/2−1
(2r +1)k
X X
= x(2r )WN2rk + x(2r + 1)WN
r =0 r =0
N/2−1 N/2−1
X X
= x(2r )(WN2 )rk + WNk x(2r + 1)(WN2 )rk
r =0 r =0

Liang Dong (Baylor University) Discrete Fourier Transform April 6, 2017 10 / 19


Decimation-in-Time Fast Fourier Transform (FFT)

2π 2π
−j N/2
WN2 = e −2j N =e = WN/2
Therefore,

N/2−1 N/2−1
X X
X (k) = x(2r )(WN2 )rk + WNk x(2r + 1)(WN2 )rk
r =0 r =0
N/2−1 N/2−1
X X
rk
= x(2r )WN/2 +WNk rk
x(2r + 1)WN/2
| r =0 {z } | r =0 {z }
G (k) H(k)

= G (k) + WNk H(k)

Liang Dong (Baylor University) Discrete Fourier Transform April 6, 2017 11 / 19


Decimation-in-Time Fast Fourier Transform (FFT)

G (k) is an (N/2)-point DFT of even samples x(2r );


H(k) is an (N/2)-point DFT of odd samples x(2r + 1).

Liang Dong (Baylor University) Discrete Fourier Transform April 6, 2017 12 / 19


Decimation-in-Time Fast Fourier Transform (FFT)

Continue...

N/4−1 N/4−1
X X
lk k lk
G (k) = g (2l)WN/4 + WN/2 g (2l + 1)WN/4
l=0 l=0
N/4−1 N/4−1
X X
lk k lk
H(k) = h(2l)WN/4 +WN/2 h(2l + 1)WN/4
| l=0 {z } | l=0 {z }
(N/4)−point DFT (N/4)−point DFT

Liang Dong (Baylor University) Discrete Fourier Transform April 6, 2017 13 / 19


Decimation-in-Time Fast Fourier Transform (FFT)

Liang Dong (Baylor University) Discrete Fourier Transform April 6, 2017 14 / 19


Decimation-in-Time Fast Fourier Transform (FFT)

Liang Dong (Baylor University) Discrete Fourier Transform April 6, 2017 15 / 19


Symmetry of WN

N/2 2π N
WN = e −j N 2 = e −jπ = −1
r +N/2 N/2
WN = WN WNr = −WNr

Liang Dong (Baylor University) Discrete Fourier Transform April 6, 2017 16 / 19


Decimation-in-Time Fast Fourier Transform (FFT)

Liang Dong (Baylor University) Discrete Fourier Transform April 6, 2017 17 / 19


Bit-reverse Reordering

x(n)’s index n binary


0 000
4 100
2 010
6 110
1 001
5 101
3 011
7 111

Liang Dong (Baylor University) Discrete Fourier Transform April 6, 2017 18 / 19


Decimation-in-Frequency Fast Fourier Transform (FFT)

Liang Dong (Baylor University) Discrete Fourier Transform April 6, 2017 19 / 19

You might also like