摘要
小波分析是分析非稳定信号的一种非常有效的方法。Mallat快速算法使得小波分析的广泛应用成为现实。在实时信号分析中 ,离散小波变换在 DSP上的有效应用受到了特别的关注。本文简单介绍了 FWT,详细阐述了在 DSP上 FWT的周期性扩展的实现。其中 ,特别介绍了 DSP的循环寻址。最后 ,给出了针对 TI公司的 TMS32 0 C3X系列 DSP的相应的汇编代码。
The wavelet transformation (WT) is a very powerful tool for the analysis of non stationary signals. The Mallat fast algorithm of WT enables the applications of wavelet transformation. An efficient implementation of the discrete form of the wavelet transform on digital signal processors (DSPs) is especially interesting for the analysis of real time signals. After a short discussion of FWT, the implementation of the periodic extension of the FWT on digital signal processors is described in detail. Within this context, special attention is given to circular addressing of digital signal processors. At last,the corresponding assembly code is described for the digital signal processor TMS320C3X series from Texas Instruments.
出处
《数据采集与处理》
EI
CSCD
2000年第2期170-175,共6页
Journal of Data Acquisition and Processing
关键词
小波变换
快速算法
数字信号处理
DSP
wavelet transformation
fast algorithm
digital signal processor(DSP)
