AMR语音编解码在DSP上的实现与优化

Implementation and Optimization of AMR Speech Codec on DSP

介绍了自适应多速率(AMR)语音编解码算法在ZSP500DSP上的代码优化。ZSP500数字信号处理器是LSILogic公司生产的第二代ZSP数字信号处理器,它是一种低功耗、高性能的DSP,广泛应用于音频、视频处理领域。简要介绍了AMR以及ZSP500的基本特性,详细讨论了基于3GPPTS26.073V6.0.0给出的C程序源代码转化为汇编在速度和空间上的优化技巧,并给出了优化结果。

This paper discusses optimized techniques for AMR speech codec algorithm on the ZSP500 DSP core. ZSP500 core is the second generation of ZSP digital signal processor architecture from LSI logic. It is a low power-consuming but high performance DSP, which is widely applied to audio and video area. This paper first introduces AMR and ZSP500 architecture, then analyzes assembly optimized techniques for both speed and space based on 3GPP TS26.073 V6.0.0 C source code. At the end, the performances are compared and some conclusions are presented.