期刊文献+

应用于时频分析的低功耗短时傅里叶变换处理器(英文)

A Low Power Short Time Fourier Transform Processor for Time-Frequency Analysis
下载PDF
导出
摘要 提出了一个应用于时频分析的短时傅里叶变换处理器。为了克服已有的离散短时傅里叶变换算法和结构的缺点,给出了一种基于快速傅里叶变换阵列的新结构。根据实际需要提出了一种新的高频域分辨率的SDF(Single-path De-lay Feedback)结构FFT单元,和传统的SDF结构FFT单元相比,反馈FIFO的深度和蝶形单元的数量都有所降低。再加上开发窗函数的对称性和适当合并硬件资源,与原始设计相比处理器的功耗降低了20%。使用中芯国际0.18微米工艺实现之后,系统工作时钟可以达到200MHz,即该处理器可以满足同样频率的采样信号的实时时频分析需求。 A discrete short time Fourier transform (STFT) processor for time - frequency analysis is proposed in this paper. To overcome the drawbacks of previous discrete STFT algorithms and architectures, a new architecture based on fast Fourier transform (FFT) array is presented. An optimized single-path delay feedback (SDF) FFT architecture is proposed for high frequency resolution, which reduces feedback FIFO depth and butterfly units' numher in comparison with the traditional one. Along with exploiting the symmetry of window function and performing hardware combination, power consumption decreased by 20%. The processor has been implemented with SMIC 0.18μm standard cell library, static timing analysis and dynamic simulation with timing back- annotated show that under worst - case conditions, 200MHz clock rate can be achieved, which means that the real - time time- frequency analysis demand can be fulfilled under the same sample frequency.
出处 《计算机技术与发展》 2008年第1期1-6,共6页 Computer Technology and Development
基金 国家863计划项目(2003AA1Z)
关键词 短时傅里叶变换 时频分析 低功耗 short time Fourier transform time- frequency analysis low power
  • 相关文献

参考文献15

  • 1Cooley J W, Tukey J W. An Algorithm for Machine Computation of Complex Fourier Series[J ]. Math Comput, 1965,19 (4) :297 - 301.
  • 2Francos A, Porat M. Non-stationary signal processing using time- frequency filter banks[ C]//Proceedings of the 13th International Conference on Digital Signal Processing. [ s. l. ] : IEEE, 1997:765 - 768.
  • 3Cao Fan, Wang Shuxun, Wang Fei. A New Time- Frequency Analysis Method of Multi - Component Chirp Signal[C].// Proceedings of the! 8th International Conference on Signal Processing. [ s. l. ] : IEEE, 2006.
  • 4Nawab S,Quatieri T,Lim J. Signal Reconstruction from Short - Time Fourier Transform Magnitude [ J ].IEEE. Transactions on Acoustics, Speech, and Signal Processing, 1983,31 (4) :986 - 998.
  • 5Goertzel G. An Algorithm for Evaluation of Finite Trigonometric Series[J]. American Mathematical Monthly, 1958,65: 34 - 35.
  • 6Farhang- Boroujeny B, Lim Y C. A Comment on Comnputational Complexity of Sliding FFT[J]. IEEE Tram on Circuits and Systems Ⅱ, 1992,39(12):875-876.
  • 7Jacobsen E, Lyons R. The Sliding DFT[J]. IEEE Signal Processing Mag, 2003, 20(2):74-80.
  • 8He Shousheng, Torkelson M. A new approach to pipeline FFT processor[C] //Proceedings of the 10th International Parallel Processing Symposium. [s. l. ]: IEEE, 1996:766 - 770.
  • 9Wold E H, Despain A M. Pipeline and parallel - pipeline FFT for VLSI impleamntation[J], IEEE Trans Computers, 1984,C- 33(5) :414 - 426.
  • 10Bi G, Jones E V. A pipelined FFT processor for word sequential data[J]. IEEE Trans Acoust, Speech, and Signal Procession, 1989,37(12) : 1982 - 1985.

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部