论文部分内容阅读
为了满足可视电话系统对高质量，低成本语音编解码器的需要，采用数字信号处理器内核（ＤＳＰＣｏｒｅ）的方法，设计了符合ＩＴＵ－ＴＧ．７２３．１建议的５．３ｋｂｉｔ／ｓ和６．３ｋｂｉｔ／ｓ双速率语音编解码器专用芯片。针对Ｇ．７２３．１算法特点，进行算法优化，如变卷积运算为加／减运算，避开求余运算等，并对需要最大运算量的情况作了模拟，保证语音编解码器在最坏情况下也能正常工作。根据专用芯片的结构，设计了编解码器的流程和接口模块，用较少的运算量和较小的存储空间，实时实现了Ｇ．７２３．１建议及其附录Ａ的全部功能和选项，并通过了全套测试序列的检验，而且留出足够多的资源实现自适应回声抵消、调制解调等其它功能。
In order to meet the requirement of the videophone system for high-quality, low-cost speech codecs, a digital signal processor core (DSPCore) method was used to design a system compliant with the ITU-TG. 723.1 Recommended 5.3kbit / s and 6.3kbit / s dual-rate voice codec chip. For G. 723.1 algorithm features, the algorithm optimization, such as the variable convolution operation for addition / subtraction operations, to avoid the arithmetic, etc., and the need for the maximum amount of computation was simulated to ensure that the voice codec in the worst case Can work normally. According to the structure of the special chip, the codec flow and interface module are designed, and G is realized in real time with less computation and less storage space. 723.1 recommendations and all the features and options in Appendix A, and passed the test of a full suite of test sequences, leaving enough resources for other functions such as adaptive echo cancellation, modulation and demodulation.