Improved Fourier and Hartley transform algorithms: application to cyclic convolution of real data

This paper highlights the possible tradeoffs between arithmetic and structural complexity when computing cyclic convolution of real data in the transform domain. Both Fourier and Hartley-based schemes are first explained in their usual form and then improved, either from the structural point of view or in the number of operations involved. Namely, we first present an algorithm for the in-place computation of the discrete Fourier transform on real data: a decimation-in-time split-radix algorithm, more compact than the previously published one. Second, we present a new fast Hartley transform algorithm with a reduced number of operations. A more regular convolution scheme based on FFT's is also proposed. Finally, we show that Hartley transforms belong to a larger class of algorithms characterized by their "generalized" convolution property.

Published in:
IEEE Transactions on Acoustics, Speech and Signal Processing, 35, 6, 818-824

 Record created 2005-04-18, last modified 2018-01-27

External link:
Download fulltext
Rate this document:

Rate this document:
(Not yet reviewed)