You are on page 1of 17

Decimation-in-time FFT Algorithms

G[0]
x[0] X[0]

N/2 POINT G[1]


x[2] X[1]

DFT G[2]
x[4] X[2]

G[3]
x[6] X[3]

WN0
Figure 9.3
H[0]
x[1] X[4]
-1
N/2 POINT H[1] WN1
x[3] X[5]
-1
H[2] WN2
x[5] X[6]
DFT -1

x[7] H[3] WN3


-1 X[7]
G1[0] G[0]
x[0] N/4point X[0]
DFT G1[1]
x[4] G[1]
X[1]

G2[0] WN0 G[2]


x[2] N/4point X[2]
DFT -1
DFT WN2
G2[1] G[3]
x[6] X[3]
-1
H1[0] H[0] WN0
x[1] N/4point
DFT X[4]
-1
DFT
H1[0] H[1] WN1
x[5] X[5]
-1
H2[0] WN0 H[2] WN2
x[3] N/4point X[6]
DFT -1 -1
DFT
H2[0] WN2 H[3] WN3
x[7]
X[7]
-1 -1

DFT
4
5
Decimation-In-Frequency FFT Algorithms
• When the output sequence X(k) is decomposed in to
smaller subsequences to compute smaller DFT’s then
that algorithm is called as decimation-in-frequency
(DIF) algorithm.
• Split the DFT equation into even and odd frequency
indexes
N −1 N / 2−1 N −1

• X 2r=  x[n]WNn2r =  x[n]WNn2r +  x[n]WNn2r


n=0 n=0 n= N / 2

N / 2−1 N −1
X 2r +1 =  x[n]W n(2r +1) +  x[n]W n (2r+1)
N /2
N/2
n=0 n= N / 2

• In the second part of summations replace n by n+N/2


N/ 2−1
(n+ N2 )2r =
 (x[n]+ x[n + N/ 2])W
N / 2−1 N / 2−1
 X 2r=  x[n]WNn2r + 
nr
x[n + 2 ]WN
N
N/ 2
n=0 n=0 n=0

8
Decimation-In-Frequency FFT Algorithm
Similarly for odd frequency indexed DFT
(n+ N2 )(2r+1)
N / 2−1 N / 2−1
 X 2r +1 =  x[n]W n(2r+1)
N +  x[n + N
2
]WN
n=0 n=0

(n+ N2 )(2r +1) n (2r+1)


=W = 1&W N = −1
N N
rN rN
WN N W WN W
N
2
N
2

(n+ N2 )(2r+1)
WN = −WNn(2r+1)
N / 2−1
 X 2r +1 =  (x[n] − x[n + N / 2])W n(2r+1)
N
n=0

(x[n] − x[n + N/ 2])W W


N / 2−1
 X 2r +1= n
N
nr
N /2
n=0

Let us define p(n) = x[n] +x[n+N / 2] and q(n) = (x[n] − x[n + N / 2])W Nn 

9
Decimation-in-frequency FFT Algorithms
14
Find DFT of a sequence x(n)={1,2,3,4,4,3,2,1}using DIF-FFT
the two kinds
of butterfly
computation
are transpose
of each other.
FFT vs. DFT
• The FFT is simply an algorithm for efficiently calculating the DFT
• Computational efficiency of an N-Point FFT:
• DFT: N2 Complex Multiplications
• FFT: (N/2) log2(N) Complex Multiplications

N DFT Multiplications FFT Multiplications FFT Efficiency

4 16 4 4:1

8 64 12 5.3:1

16 256 32 8:1

32 1024 80 12.8:1

64 4096 192 21.3:1

128 16384 448 36.5:1

256 65,536 1,024 64 : 1

512 262,144 2,304 114 : 1

1,024 1,048,576 5,120 205 : 1

You might also like