采用预测残差的回声消除单双端检测算法
|
摘要
在单端与双端状态下,自适应滤波器的预测残差数值迥异,对预测残差数值进行归一化处理,拉开单端、双端状态的数值差距,扩大阈值范围.提出基于预测残差的回声消除单双端判断方法,并提出在单双端交界处使用自适应滤波器权值复用与更新机制,以及算法的实现步骤.实验结果表明:该算法具有宽广的阈值选择范围,可以有效防止因单双端误判引起的滤波器发散,减小总的回声消除预测误差.
In the condition of single-talk and double-talk,the predictive residual data of self-adapting filter is different.this article set normalization processing in predictive residual data,opened numerical gap in singletalk and double-talk,and expanded the scope of the threshold value,then this article put forward the echo cancellation single and double-talk judgment method based on predictive residuals.And put forward that it can use the adaptive filter working mechanism of weight reuse and update in single-talk and double-talk junction.At last,the implementation steps of the algorithm are also given.The experimental results showed that this algorithm has a wide range of threshold selection,it can effectively prevent the filter divergence caused by inaccurate double-talk detection and reduce the total echo cancellation prediction error.
In the condition of single-talk and double-talk,the predictive residual data of self-adapting filter is different.this article set normalization processing in predictive residual data,opened numerical gap in singletalk and double-talk,and expanded the scope of the threshold value,then this article put forward the echo cancellation single and double-talk judgment method based on predictive residuals.And put forward that it can use the adaptive filter working mechanism of weight reuse and update in single-talk and double-talk junction.At last,the implementation steps of the algorithm are also given.The experimental results showed that this algorithm has a wide range of threshold selection,it can effectively prevent the filter divergence caused by inaccurate double-talk detection and reduce the total echo cancellation prediction error.
引文
[1]西蒙.赫金.自适应滤波器原理[M].北京:电子工业出版社,2003:251-270.
[2]宋知用.MATLAB在语音信号分析与合成中的应用[M].北京:北京航空航天大学出版社,2013:284-320.
[3]NEKUII M,ATARODI M.Composite PNLMS&NLMS adaptation:A new method for network echo cancellation[C]∥14th International Conference on Digital Signal Processing.Santorini:IEEE Press,2002:757-760.
[4]DENG H,DOROSLOVACKI M.Improving convergence of the PNLMS algorithm for sparse impulse response identification[J].Signal Processing Letters,2005,12(3):181-184.
[5]刘立刚,FUKUMOTO M,张世永.一种变步长Proportionate NLMS自适应滤波算法及其在网络回声消除中的应用[J].电子学报,2010,38(4):973-978.
[6]尹浩琼.通信系统中回声消除的研究[D].北京:北京邮电大学,2006:20-35.
[7]江冰,叶玲,朱成健,等.一种提高稳态误差及收敛速度的回声消除算法[J].声学技术,2009,28(3):295-299.
[8]YE Hua,WU Boxiu.A new double-talk detection algorithm based on the orthogonality theorem[J].IEEE Transactions on Communications,1991,39(11):1542-1545.
[9]AHGREN P,JAKOBSSON A.A study of double-talk detection performance in the presence of acoustic echo path changes[J].IEEE Transactions on Consumer Electronics,2006,52(2):515-522.
[10]BENESTY J,MORGAN D R,CHO J H.A new class of doubletalk detectors based on cross-correlation[J].IEEE Transactions on Speech and Audio Processing,2000,8(2):168-172.
[11]宋彦,汪萌,戴礼荣.带双端发声检测的回声消除改进算法[J].清华大学学报(自然科学版),2008(增刊1):760-764.
[12]许礼武,张伟彬,陈波.一种新的回音消除的双向通话检测算法[J].声学技术,2008,27(4):539-542.
[13]吴超,付强,颜永红.基于噪声估计和能量比的双讲检测方法[C]∥第十二届全国人机语音通讯学术会议.贵阳:[出版者不详],2013:5.
[14]谢鹏,刘加.采用变步长算法的无双端检测回声消除系统[J].数据采集与处理,2010,25(6):736-740.
[15]万政伟,惠晓威,朱婷.基于NLMS算法的自适应滤波器的研究与应用[J].计算机系统应用,2010,19(4):46-48.
[16]范团宝.基于滑动窗增强NLMS的回声消除算法研究[C]∥2011年度全国无线及移动通信学术大会.北京:[出版者不详],2011:7.
[17]International Telecommunication Union.ITU-T recommendation G.168[S].Geneva:ITU,2000:86-105.
[2]宋知用.MATLAB在语音信号分析与合成中的应用[M].北京:北京航空航天大学出版社,2013:284-320.
[3]NEKUII M,ATARODI M.Composite PNLMS&NLMS adaptation:A new method for network echo cancellation[C]∥14th International Conference on Digital Signal Processing.Santorini:IEEE Press,2002:757-760.
[4]DENG H,DOROSLOVACKI M.Improving convergence of the PNLMS algorithm for sparse impulse response identification[J].Signal Processing Letters,2005,12(3):181-184.
[5]刘立刚,FUKUMOTO M,张世永.一种变步长Proportionate NLMS自适应滤波算法及其在网络回声消除中的应用[J].电子学报,2010,38(4):973-978.
[6]尹浩琼.通信系统中回声消除的研究[D].北京:北京邮电大学,2006:20-35.
[7]江冰,叶玲,朱成健,等.一种提高稳态误差及收敛速度的回声消除算法[J].声学技术,2009,28(3):295-299.
[8]YE Hua,WU Boxiu.A new double-talk detection algorithm based on the orthogonality theorem[J].IEEE Transactions on Communications,1991,39(11):1542-1545.
[9]AHGREN P,JAKOBSSON A.A study of double-talk detection performance in the presence of acoustic echo path changes[J].IEEE Transactions on Consumer Electronics,2006,52(2):515-522.
[10]BENESTY J,MORGAN D R,CHO J H.A new class of doubletalk detectors based on cross-correlation[J].IEEE Transactions on Speech and Audio Processing,2000,8(2):168-172.
[11]宋彦,汪萌,戴礼荣.带双端发声检测的回声消除改进算法[J].清华大学学报(自然科学版),2008(增刊1):760-764.
[12]许礼武,张伟彬,陈波.一种新的回音消除的双向通话检测算法[J].声学技术,2008,27(4):539-542.
[13]吴超,付强,颜永红.基于噪声估计和能量比的双讲检测方法[C]∥第十二届全国人机语音通讯学术会议.贵阳:[出版者不详],2013:5.
[14]谢鹏,刘加.采用变步长算法的无双端检测回声消除系统[J].数据采集与处理,2010,25(6):736-740.
[15]万政伟,惠晓威,朱婷.基于NLMS算法的自适应滤波器的研究与应用[J].计算机系统应用,2010,19(4):46-48.
[16]范团宝.基于滑动窗增强NLMS的回声消除算法研究[C]∥2011年度全国无线及移动通信学术大会.北京:[出版者不详],2011:7.
[17]International Telecommunication Union.ITU-T recommendation G.168[S].Geneva:ITU,2000:86-105.