Decimation in frequency The radix-2 decimation-in-frequency algorithm rearranges the discrete Fourier transform (DFT) equation into two parts: computation of the even-numbered discrete-frequency indices X k for k 0 2 4 … N 2 (or X 2 r as in ) and computation of the odd-numbered indices k 1 3 5 … N 1 (or X 2 r 1 as in ) X 2 r n N 1 0 x n W N 2 r n n N 2 1 0 x n W N 2 r n n N 2 1 0 x n N 2 W N 2 r ( n + N 2 ) n N 2 1 0 x n W N 2 r n n N 2 1 0 x n N 2 W N 2 r n 1 n N 2 1 0 x n x n N 2 W N 2 r n DFT N 2 x n x n N 2 X 2 r 1 n N 1 0 x n W N ( 2 r + 1 ) n n N 2 1 0 x n W N N 2 x n N 2 W N ( 2 r + 1 ) n n N 2 1 0 x n x n N 2 W N n W N 2 r n DFT N 2 x n x n N 2 W N n The mathematical simplifications in and reveal that both the even-indexed and odd-indexed frequency outputs X k can each be computed by a length- N 2 DFT. The inputs to these DFTs are sums or differences of the first and second halves of the input signal, respectively, where the input to the short DFT producing the odd-indexed frequencies is multiplied by a so-called twiddle factor term W N k 2 k N . This is called a decimation in frequency because the frequency samples are computed separately in alternating groups, and a radix-2 algorithm because there are two groups. graphically illustrates this form of the DFT computation. This conversion of the full DFT into a series of shorter DFTs with a simple preprocessing step gives the decimation-in-frequency FFT its computational savings.

Decimation in frequency of a length-N DFT into two length-N2 DFTs preceded by a preprocessing stage. Whereas direct computation of all N DFT frequencies according to the DFT equation would require N 2 complex multiplies and N 2 N complex additions (for complex-valued data), by breaking the computation into two short-length DFTs with some preliminary combining of the data, as illustrated in , the computational cost is now New Operation Counts 2 N 2 2 N N 2 2 N 2 complex multiplies 2 N 2 N 2 1 N N 2 2 complex additions This simple manipulation has reduced the total computational cost of the DFT by almost a factor of two!The initial combining operations for both short-length DFTs involve parallel groups of two time samples, x n and x n N 2 . One of these so-called butterfly operations is illustrated in . There are N 2 butterflies per stage, each requiring a complex addition and subtraction followed by one twiddle-factor multiplication by W N n 2 n N on the lower output branch.

DIF butterfly: twiddle factor after length-2 DFT It is worthwhile to note that the initial add/subtract part of the DIF butterfly is actually a length-2 DFT! The theory of multi-dimensional index maps shows that this must be the case, and that FFTs of any factorable length may consist of successive stages of shorter-length FFTs with twiddle-factor multiplications in between. It is also worth noting that this butterfly differs from the decimation-in-time radix-2 butterfly in that the twiddle factor multiplication occurs after the combining.

Radix-2 decimation-in-frequency algorithm The same radix-2 decimation in frequency can be applied recursively to the two length- N 2 DFTs to save additional computation. When successively applied until the shorter and shorter DFTs reach length-2, the result is the radix-2 decimation-in-frequency FFT algorithm.

Radix-2 decimation-in-frequency FFT for a length-8 signal The full radix-2 decimation-in-frequency decomposition illustrated in requires M 2 N stages, each with N 2 butterflies per stage. Each butterfly requires 1 complex multiply and 2 adds per butterfly. The total cost of the algorithm is thus Computational cost of radix-2 DIF FFT N 2 2 N complex multiplies N 2 N complex adds This is a remarkable savings over direct computation of the DFT. For example, a length-1024 DFT would require 1048576 complex multiplications and 1047552 complex additions with direct computation, but only 5120 complex multiplications and 10240 complex additions using the radix-2 FFT, a savings by a factor of 100 or more. The relative savings increase with longer FFT lengths, and are less for shorter lengths. Modest additional reductions in computation can be achieved by noting that certain twiddle factors, namely W N 0 , W N N 2 , W N N 4 , W N N 8 , W N 3 N 8 , require no multiplications, or fewer real multiplies than other ones. By implementing special butterflies for these twiddle factors as discussed in FFT algorithm and programming tricks, the computational cost of the radix-2 decimation-in-frequency FFT can be reduced to 2 N 2 N 7 N 12 real multiplies 3 N 2 N 3 N 4 real additions The decimation-in-frequency FFT is a flow-graph reversal of the decimation-in-time FFT: it has the same twiddle factors (in reverse pattern) and the same operation counts.In a decimation-in-frequency radix-2 FFT as illustrated in , the output is in bit-reversed order (hence "decimation-in-frequency"). That is, if the frequency-sample index n is written as a binary number, the order is that binary number reversed. The bit-reversal process is illustrated here.It is important to note that, if the input data are in order before beginning the FFT computations, the outputs of each butterfly throughout the computation can be placed in the same memory locations from which the inputs were fetched, resulting in an in-place algorithm that requires no extra memory to perform the FFT. Most FFT implementations are in-place, and overwrite the input data with the intermediate values and finally the output.