ResearchBib Share Your Research, Maximize Your Social Impacts
Sign for Notice Everyday Sign up >> Login

ON THE COMPARATIVE ANALYSIS OF THE FAST BLUESTEIN AND FAST COOLEY-TUKEY NUMERICAL ALGORITHMS FOR DIGITAL AND ANALOG SIGNALS PROCESSING

Journal: INTERNATIONAL JOURNAL OF RESEARCH -GRANTHAALAYAH (Vol.3, No. 11)

Publication Date:

Authors : ; ; ; ;

Page : 133-146

Keywords : Algorithm; FFT; Cooley-Tukey; Bluestein; Comparasion; Analisis.;

Source : Downloadexternal Find it from : Google Scholarexternal

Abstract

This study was designed to compare the computing efficiency of FC-TNADSP and the FBNADSP.to ascertain a faster numerical algorithm necessary for the processing of digital signals. The faster numerical algorithm established in this study is abbreviated with RCFCTNADSP (Compared Resultant-TNADSP). The methodology adopted in this work was comparative analysis development design. The major technologies used in this work are the FCTNADSP and FBNADSP which were hitherto simulated on the c++ programming technologies. The c++ served as a signal processing language simulator (SPLS). The execution times of the fast Cooley-Tukey and the fast Bluestein algorithms were 1.70 seconds and 1.74 seconds respectively. On comparing the speeds of the fast Cooley–Tukey and the fast Bluestein algorithms we observed that the fast Cooley-Tukey algorithm has 0.04 seconds speed improvement over the fast Bluestein algorithm. In line with this outcome, we concluded that the fast Cooley-Tukey algorithm (FC-TNADSP) is faster than the fast Bluestein algorithm (FBNADSP). In the same vein the fast Cooley-Tukey algorithm (FC-TNADSP algodsp-2) is therefore the fastest DSP algorithm. This is however faster than the spectrum of FFT algorithms of O(nlogn) computing speed, a speed considered to be the fastest hitherto. The result of this study shows we can have faster numerical algorithms other than the traditional spectrum of FFT algorithms of O(nlogn) computing speed. The algorithms were tested on input block of width 1000 units, and above, and can be implemented on input size of 100 000, and 1000 000 000 without the challenge of storage overflow. The input samples tested in this work was the discretized pulse wave form with undulating shape out of which the binary equivalents were extracted. Other forms of signals may also be tested in this fast algorithm provided they are interpreted in the digital wave type.

Last modified: 2017-09-24 17:25:37