透過您的圖書館登入
IP:13.58.60.192
  • 學位論文

利用雙核心提升快速傅立葉轉換效能

Application of Multi-core to Improve the Computation Efficiency of Fast Fourier Transform

指導教授 : 謝承道

摘要


由於電腦軟硬體設備在近期內發展非常迅速,所以在市面上雙/多核心處理器已非常普及。但在處理單工任務方面,雙/多核心會因為多數軟體不支援,導致實際上真正在工作的只有一顆核心,其餘核心則閒置,這對其效能便是一種浪費。所以本文主要是使用軟體,使得閒置核心在對於處理單工方面也能夠工作來達到提昇效能作用。 本文是藉由開啟雙核心功能來提升快速傅立葉轉換效能,於Windows XP作業系統上使用軟體C++內中的OpenMP來達到開啟雙核心功能,在套用至快速傅立葉運算中,來達到確切的效能提升。在實驗驗證方面,則藉以有無開啟雙核心功能來觀察其運算效能變化。實驗結果顯示,開啟雙核心功能之後,若資料筆數夠大,則快速傅立葉轉換之計算時間可大幅縮減,以驗證本方法之可行性。

並列摘要


Dual / multi-core processors are already in common used as a result of the extraordinarily rapid development of computer hardware and software equipments in recent years. Most software not supporting the dual / multi-core workers in dealing with single-task leads to it works with only one core, and the remaining cores are idle. It is a waste of efficiency. In this thesis, total cores are used for a single-task in order to improve the efficiency. The computation efficiency of Fast Fourier Transform is enhanced by using the dual-core function in this thesis. The dual-core function is enabled by OpenMP in C++, and then applied to the computation of Fast Fourier Transform to improve the performance. In the experiments, the different in computation time between with and without using the dual-core function is observed. The test results exhibit that if the number of data is enough the computation time can be reduced significantly. The feasibility of the approach is verified.

並列關鍵字

Dual-core Fast Fourier Transform

參考文獻


[6] 林文祥,“應用平行運算於小波轉換分析電力品質信號”,碩士論文,崑山科技大學,2008年。
[7] 郭吉明,“應用分散式平行計算於諧波分析系統之建立”,碩士論文,崑山科技大學,2008年。
[2] 蔡坤成,“電力品質訊號之頻譜分析及探討”, 碩士論文,崑山科技大學,2005年。
[15] Smotherman, Mark(2005).History of Multithreading.
參考文獻

被引用紀錄


許永明(2011)。利用多核心技術提升快速傅立葉轉換效能〔碩士論文,崑山科技大學〕。華藝線上圖書館。https://doi.org/10.6828/KSU.2011.00080

延伸閱讀