A novel approach for FFT data reordering
Marwan A. Jaber, Daniel Massicotte · 2010
The Fast Fourier Transform (FFT) is a key role in signal processing applications that is useful for the frequency domain analysis of signals. The FFT computation requires an indexing scheme at each stage to address input/output data and coefficient multipliers properly. Most of these indexing schemes are based on bit-reversal techniques that are boosted by a look-up table requiring extra memory storage. This paper describes a novel data reordering technique based on the vector calculation of size r. FFTs are considered in-place (or in situ) algorithms that transform a data structure by using a constant amount of memory storage. We demonstrate that our proposed method reduces memory usage by eliminating the look-up table traditionally employed in the computation of bit-reversal indexes.