A comment on the computational complexity of sliding FFT
- 1 December 1992
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems II: Analog and Digital Signal Processing
- Vol. 39 (12), 875-876
- https://doi.org/10.1109/82.208583
Abstract
The sliding fast Fourier transform (FFT) is reviewed and is shown to have the computational complexity of N complex multiplications per sample, as opposed to the well-cited assumption of (N/2) log/sub 2/ N complex multiplication per sample reported in a book by L.R. Rabiner and B. Gold (1975).<>Keywords
This publication has 7 references indexed in Scilit:
- The use of orthogonal transforms for improving performance of adaptive filtersIEEE Transactions on Circuits and Systems, 1989
- Transform domain LMS algorithmIEEE Transactions on Acoustics, Speech, and Signal Processing, 1983
- Generalized running discrete transformsIEEE Transactions on Acoustics, Speech, and Signal Processing, 1982
- Adaptive frequency sampling filtersIEEE Transactions on Acoustics, Speech, and Signal Processing, 1981
- A continuous recursive DFT analyzer--The discrete coherent memory filterIEEE Transactions on Acoustics, Speech, and Signal Processing, 1980
- Recursive discrete Fourier transformationIEEE Transactions on Acoustics, Speech, and Signal Processing, 1980
- Recursive and nonrecursive realizations of digital filters designed by frequency sampling techniquesIEEE Transactions on Audio and Electroacoustics, 1971