{"title":"Double factors algorithm for computing DFT","authors":"Haijun Li, Caojun Yan, Wenbiao Peng","doi":"10.1109/IASP.2009.5054641","DOIUrl":null,"url":null,"abstract":"A fast Fourier transform algorithm for computing N=N<inf>1</inf>×N<inf>2</inf>-point DFT, where both factors N<inf>1</inf> and N<inf>2</inf> are smaller positive integer, said to be a double factors algorithm(DFA), is developed. The DFA subdivides a DFT of length N=N<inf>1</inf>×N<inf>2</inf> into smaller transforms of length N<inf>1</inf> and N<inf>2</inf> and takes the following steps:(1) computes N<inf>1</inf> N<inf>2</inf>-point DFTs , (2) multiplies the values of DFT by twiddle factors, (3) computes N<inf>2</inf> N<inf>1</inf>-point DFTs. The structure of the DFA is similar to those of the most simple PFA and WFTA, but N<inf>1</inf> and N<inf>2</inf> are not necessarily relatively prime. When N=2<sup>M</sup> or 4<sup>M</sup>, the total number of computations of DFT in the DFA is less than those in the radix-2 and radix-4 FFT algorithm but slightly more than that in the split-radix FFT algorithm. When N is other values, the total number of computations of DFT in the DFA is less than those in the PFA and WFTA.","PeriodicalId":143959,"journal":{"name":"2009 International Conference on Image Analysis and Signal Processing","volume":"19 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2009-04-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"4","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2009 International Conference on Image Analysis and Signal Processing","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/IASP.2009.5054641","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 4
Abstract
A fast Fourier transform algorithm for computing N=N1×N2-point DFT, where both factors N1 and N2 are smaller positive integer, said to be a double factors algorithm(DFA), is developed. The DFA subdivides a DFT of length N=N1×N2 into smaller transforms of length N1 and N2 and takes the following steps:(1) computes N1 N2-point DFTs , (2) multiplies the values of DFT by twiddle factors, (3) computes N2 N1-point DFTs. The structure of the DFA is similar to those of the most simple PFA and WFTA, but N1 and N2 are not necessarily relatively prime. When N=2M or 4M, the total number of computations of DFT in the DFA is less than those in the radix-2 and radix-4 FFT algorithm but slightly more than that in the split-radix FFT algorithm. When N is other values, the total number of computations of DFT in the DFA is less than those in the PFA and WFTA.