Performance analysis of short-time fourier transform algorithm with selected optimization methods

The paper consists of well-known Fourier Transform algorithms review with its optimization methods. Beyond others, root of unity lookup table, in-situ computing, bit-reversal permutation, feed forward method and metaprogramming approaches can be pointed out. The main aim of the presented work was to...

Full description

Saved in:

Bibliographic Details
Published in:	2018 International Interdisciplinary PhD Workshop (IIPhDW) pp. 371 - 376
Main Authors:	Rybak, Grzegorz, Chaniecki, Zbigniew, Grudzien, Krzysztof, Romanowski, Andrzej, Sankowski, Dominik
Format:	Conference Proceeding
Language:	English
Published:	IEEE 01-05-2018
Subjects:	Arrays bit-reversal permutation Discrete Fourier transforms ECT feedforward Feeds Indexes Iterative methods meta-programming Microsoft Windows optimisation Short-time Fourier transform STFT
Online Access:	Get full text
Tags:	Add Tag No Tags, Be the first to tag this record!

Description
Summary:	The paper consists of well-known Fourier Transform algorithms review with its optimization methods. Beyond others, root of unity lookup table, in-situ computing, bit-reversal permutation, feed forward method and metaprogramming approaches can be pointed out. The main aim of the presented work was to increase currently achieved Short-Time Fourier Transform (STFT) performance using Java programming language. Authors present a new iterative approach. The idea of the solution was to reduce unnecessary operations, by extracting computation and maintenance of loops' indices and twiddle factors' computation as well as reordering output data to the meta-program execution phase, in order to gain strict runtime execution. Presented methods are successfully adapted to electrical capacitance tomography images processing.
DOI:	10.1109/IIPHDW.2018.8388393