基于DSP的语音编码算法及其实现

【作者】 王锋；

【导师】 廉保旺；

【作者基本信息】 西北工业大学 ， 通信与信息系统， 2001， 硕士

【摘要】 本文在分析了由语音信号的产生机理所得出的短时平稳特性的基础上，利用线性预测编码（LPC）技术，分析了码激励线性预测（CELP）编码的基本原理。通过对ITU-T G.729共轭结构-代数码激励线性预测（CS-ACELP）算法的分析，提出了基于高级DSP TMS320VC5402硬件系统的核算法的实现方案。文中给出了语音信号压缩编码的具体实现方法，并对其进行了验证。结果表明，该系统具有功耗低、实用性强、运算迅速而准确等显著特点。本文对多媒体通信系统中的语音传输有一定的参考价值。更多还原

【Abstract】 Based on the analysis of short-time calm characteristics of audio signal educing from produce-mechanism of audio signal , and the Linear Prediction coding technology , the basic principle of Code-Excited Linear-Prediction coding is discussed . After the algorithm of audio compression coding which named Conjugate Structure Algebraic Code Excited Linear Prediction (CS-ACELP) and described in the ITU-T Recommendation G.729 is analyzed, its implement scheme founded on the hardware system composed of advanced DSP TMS320VC5402 is presented The specific operation method about compressing and coding audio signal is presented , and is validated . Experiments prove that the system has some characteristics such as low power consumption , practicability , rapid and exact calculate , and so on.更多还原