S. K. Shome, Abhinav Ahesh, Durgesh Kr Gupta, Srk Vadali
{"title":"具有高效寻址逻辑的高度可编程的Radix-2 FFT处理器的体系结构设计","authors":"S. K. Shome, Abhinav Ahesh, Durgesh Kr Gupta, Srk Vadali","doi":"10.1109/ICDCSYST.2012.6188802","DOIUrl":null,"url":null,"abstract":"A large number of efficient fixed geometry Fast Fourier Transform (FFT) VLSI designs have been developed till date. We propose a novel architectural design for a highly programmable Radix-2 Decimation-In-Frequency (DIF) FFT processor using relatively simple memory addressing logic. The 5-level programmability of the design, allows computation of 64, 128, 256, 512 or 1024 point FFT of the input signal, depending on application. Besides, the architecture provides the flexibility of computing an N point FFT for M length data (N >; M), i.e. with an enhanced resolution also. A complete system flow of the entire FFT architecture along with twiddle factor multiplication, bit reversal and a detailed efficient Address Generation Block (AGB) are also presented. The address generation methodology adopted for the proposed design is based on counters and multiplexers which significantly saves the hardware as well as the latency requirement introduced thereon.","PeriodicalId":356188,"journal":{"name":"2012 International Conference on Devices, Circuits and Systems (ICDCS)","volume":"78 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2012-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"9","resultStr":"{\"title\":\"Architectural design of a highly programmable Radix-2 FFT processor with efficient addressing logic\",\"authors\":\"S. K. Shome, Abhinav Ahesh, Durgesh Kr Gupta, Srk Vadali\",\"doi\":\"10.1109/ICDCSYST.2012.6188802\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"A large number of efficient fixed geometry Fast Fourier Transform (FFT) VLSI designs have been developed till date. We propose a novel architectural design for a highly programmable Radix-2 Decimation-In-Frequency (DIF) FFT processor using relatively simple memory addressing logic. The 5-level programmability of the design, allows computation of 64, 128, 256, 512 or 1024 point FFT of the input signal, depending on application. Besides, the architecture provides the flexibility of computing an N point FFT for M length data (N >; M), i.e. with an enhanced resolution also. A complete system flow of the entire FFT architecture along with twiddle factor multiplication, bit reversal and a detailed efficient Address Generation Block (AGB) are also presented. The address generation methodology adopted for the proposed design is based on counters and multiplexers which significantly saves the hardware as well as the latency requirement introduced thereon.\",\"PeriodicalId\":356188,\"journal\":{\"name\":\"2012 International Conference on Devices, Circuits and Systems (ICDCS)\",\"volume\":\"78 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2012-03-15\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"9\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"2012 International Conference on Devices, Circuits and Systems (ICDCS)\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1109/ICDCSYST.2012.6188802\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"2012 International Conference on Devices, Circuits and Systems (ICDCS)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ICDCSYST.2012.6188802","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
Architectural design of a highly programmable Radix-2 FFT processor with efficient addressing logic
A large number of efficient fixed geometry Fast Fourier Transform (FFT) VLSI designs have been developed till date. We propose a novel architectural design for a highly programmable Radix-2 Decimation-In-Frequency (DIF) FFT processor using relatively simple memory addressing logic. The 5-level programmability of the design, allows computation of 64, 128, 256, 512 or 1024 point FFT of the input signal, depending on application. Besides, the architecture provides the flexibility of computing an N point FFT for M length data (N >; M), i.e. with an enhanced resolution also. A complete system flow of the entire FFT architecture along with twiddle factor multiplication, bit reversal and a detailed efficient Address Generation Block (AGB) are also presented. The address generation methodology adopted for the proposed design is based on counters and multiplexers which significantly saves the hardware as well as the latency requirement introduced thereon.