题目：G.729语音编解码算法及其在DSP上的实现（包括PPT演示等）

　中 文 摘 要 多媒体技术的不断发展，对语音编码提出的要求也越来越高。不仅要求低码率、低延时，而且要求很高的语音质量。而作为中低速率语音编码算法中的佼佼者G.729越来越受到广泛的重视。 ITU-T G.729语音编码方案采用共轭结构代数码本激励线性预测技术（CS-ACELP），算法具有8Kbit/s的编码速率、较低的延迟和高编码语音质量。它在会议电视、数字综合业务网、网络多媒体通信和语音存储系统等领域得到广泛的应用。近年来，高速数字信号处理器的发展使G.729这类高复杂的编解码器的实现成了可能。 　本文在介绍了语音编码原理的基础上，详细分析了CS－ACELP语音编码算法原理及其特性。运用标准C语音按照结构化设计方法实现了算法的仿真，并用标准的测试序列对仿真程序进行了测试，验证了其正确性。 在此基础上，基于TI　TMS320C6711　DSP硬件平台，实现了对语音信号的编解码。合成语音信号经过主观测试，语音质量良好，满足通信要求。 关键词：语音编码 语音解码 CS－ACELP G.729 DSP Abstract With the improvement of the multimedia communication technology, common users now ask for more and more convenience for speech coding. Not only bit rate and low delay but also high quality speech coding schemes are needed. As a good arithmetic of low speed speech coding arithmetic, G729 is regarded extensively. ITU-T G.729 speech coding protocol , which adopted the Conjungate-Structure Algebraic-Code-Excited Linear-Prediction (CS-ACELP)technology. It has the performance of coding rate of 8Kbit/s, low delay and high reconstructed speech quality. It is widely used in video telephony, B-ISDN, network multimedia systems and digital speech storage systems, etc. Due to the development of DSPs chips, it is now possible for us to process the speech signals with complicated algorithms such as G.729. In this thesis, after introducing the principle of speech coding, the principle and characteristic of CS-ACELP are analysed in detail. With the computer language ANSI-C employed, CS-ACELP is realized using the structured design method. Validity of the simulator is justified with the standard test vectors. And then, speech coding and decoding are realized at the base of TI TMS320C6711 DSP chip. The reconstructed speech is proved high quality , and meet the requirement in telecommunication. Key words: Speech coding Speech decoding CS-ACELP G.729 DSP 目 录 中文摘要..................................................................1 ABSTRACT...............................................................2 第一章　引言..............................................................1 1.1　课题来源与选题依据.................................................1 1.2　课题研究的意义与价值...............................................1 1.3　研究内容与论文结构.................................................2 第二章　语音编码理论基础..................................................2 2.1　语音信号的产生模型与特点...........................................3 2.2　语音编码的分类与比较...............................................4 2.3　语音编码中常用的处理技术...........................................5 　　2.3.1 短时分析技术....................................................5　 2.3.2 线性预测编码（LPC）分析.........................................6　 2.3.3 矢量量化........................................................6　 2.3.4 码本激励线性预测................................................7　第三章　CS－ACELP语音编解码算法分析....................................8 3.1　CS－ACELP算法概述..................................................8 3.2　CS－ACELP语音编码原理.............................................10 　　3.2.1 预处理.........................................................10 3.2.2 LP分析、量化、内插.............................................10 3.2.3 感知加权........................................................12　 3.2.4 基音分析........................................................13　 3.2.5 固定码本结构及其搜索............................................15　 3.2.6 增益量化........................................................16 3.3　CS－ACELP语音解码原理..............................................17 　　3.3.1 参数解码........................................................17　 3.3.2 重构语音信号....................................................18　 3.3.3 后置滤波器......................................................18　 3.3.4 错误帧屏蔽处理..................................................20　第四章　TMS320C6711开发系统............................................20 4.1　TMS3206711DSK硬件平台..............................................21 4.2　DSP软件开发平台....................................................22 第五章　G.729在TMS320C6711开发系统上的实现.............................24 5.1　G.729标准算法C程序的实现............................