语音声码器模型及ADPCM声码器仿真  

The Model of Speech Codec and the Simulation of ADPCM's Codec

作　　者：刘华[1] 

机构地区：[1]南华大学电气工程学院,湖南衡阳421001

出　　处：《长春师范学院学报（自然科学版）》2013年第3期22-24,共3页Journal of Changchun Teachers College

基　　金：湖南省教育厅科研项目(09C855)

摘　　要：ADPCM是一种波形编码,对应的标准算法有G.721,G.726等。本文建立形式化的数学模型,分析编解码算法及编码前后、解码前后信息差度量Δ1、Δ2。对0.8411秒的WAV语音实现了采样频率8KHz,数据率32Kbps的语音编码,并很好地还原出声音。实现算法时要充分考虑环境噪音。算法对[-0.1,0.1]的随机加性噪声有较好的鲁棒性,编解码时间稳定,时间复杂度为O(N)。ADPCM is a kind of the wave coding.The algorithms matching ADPCM are G.721,G.726 and so on.The important parameters such as the difference of information Δ1 andΔ2 are analyzed by constructing formulated mathematics model.The length of wave file is 0.8411 second,the sampling frequency is 8 KHz,the code rate is 32Kbps,and the quality of sound after decoder is very good.The process about realizing algorithm takes the environment＇s noise into consideration.The algorithm has the good robustness with the noise in the scale [-0.1,0.1].The time is stable and the complexity is O（N）.

关 键 词：声码器 自适应差分脉冲编码调制 数学模型 

分 类 号：TN912[电子电信—通信与信息系统]