摘要
自适应多速率宽带语音编码标准(AMR-WB)是第一个同时用于有线和无线通用标准的宽带语音编码器。介绍了AMR-WB算法并进行了AMR-WB算法定点C代码的仿真实现以及将该算法移植到TMS320VC5509DSP芯片的实现过程,具体包括C5509存储器的分配,C5509与PC机兼容性问题的处理;并对该算法进行初步优化,基本算子的优化和重点函数的优化;搭建了系统的硬件框架并设计了系统的软件结构,实时实现了该算法;最后进行了测试和评价。
Adaptive Multi-Rate Wideband speech coding standard (AMR-WB) is the first wideband speech codec used for both wired and wireless universal standards. First, the AMR-WB algorithm is introduced, and the fixed- point C code is simulated. The process of AMR-WB algorithm transplanted to TMS320VC5509 DSP chip is introduced, including the memory allocation of C5509, solving the compatibility problem between C5509 and PC machine. Then the preliminary optimization, the basic operator optimization and the optimization of key functions of the algorithm are completed. And then the framework of system hardware is constructed and the system software architecture is designed, and real-time implementation of the algorithm is achieved. Finally, the system is tested and evaluated.
出处
《电声技术》
2011年第1期67-71,共5页
Audio Engineering
基金
山西省自然科学基金项目(20051039)
2009年太原理工大学校科技发展基金项目