The additive FFT algorithms of the preceding two chapters make no explicit use of the multiplicative structure of the indexing set. We will see how the multiplicative structure can be applied, in the case of transform size N = RS, where R and S are relatively prime, to design an FT algorithm that is similar in structure to these additive algorithms but no longer requires the twiddle factor multiplication. The idea is due to Good  in 1958 and Thomas  in 1963, and the resulting algorithm is called the Good-Thomas Prime Factor algorithm (PFA).
KeywordsDiscrete Fourier Transform Permutation Matrix Index Point Permutation Matrice Multiplicative Structure
Unable to display preview. Download preview PDF.
- Kolba, D. P. and Parks, T. W. “A Prime Factor FFT Algorithm Using high-speed Convolution”, IEEE Trans. Acoust., Speech and Signal Proc., 25, 1977.Google Scholar
- Temperton, C. “A New Set of Minimum-add Small-n Rotated DFT Modules”, to appear in J. Comput. Phys. Google Scholar
- Temperton, C. “Implementation of A Prime Factor FFT Algorithm on CRAY-1”, to appear in Parallel Computing. Google Scholar
- Temperton, C. “A Self-Sorting In-place Prime Factor Real/half-complex FFT Algorithm”, to appear in J. Comput. Phys. Google Scholar
- Thomas, L. H. “Using a Computer to Solve Problems in Physics”, in Applications of Digital Computers, Ginn and Co., 1963.Google Scholar