变阶变正则化APA在回声抵消中的应用

计算机工程与应用 ›› 2014, Vol. 50 ›› Issue (15): 168-171.

变阶变正则化APA在回声抵消中的应用

冀常鹏1，2，姬红红1，郭伟平1

1.辽宁工程技术大学研究生院，辽宁葫芦岛 125105
2.辽宁工程技术大学电子与信息工程学院，辽宁葫芦岛 125105

出版日期:2014-08-01 发布日期:2014-08-04

APA with evolving order and variable regularization for echo cancellation

JI Changpeng1，2, JI Honghong1, GUO Weiping1

1.Institute of Graduate, Liaoning Technical University, Huludao, Liaoning 125105, China
2.School of Electronic and Information Engineering, Liaoning Technical University, Huludao, Liaoning 125105, China

Online:2014-08-01 Published:2014-08-04

摘要/Abstract

摘要： 目前仿射投影算法已成为回声抵消研究中常用的算法。针对收敛速度和稳态误差这对相互矛盾因素，提出了变正则化因子和变投影阶数相结合的新算法。通过变正则化因子和可变投影阶数的协同调节，使收敛速度和稳态误差达到相对较好的状态。摒弃后验误差为零的传统思想，将噪声的统计特性考虑到自适应过程中。推导出了算法具体的更新公式。实验结果表明改进的算法要比变阶数APA或变正则化因子APA具有较小的稳态误差和较快的收敛速度。

关键词: 变正则化因子, 变阶数, 仿射投影算法（APA）, 声学回声抵消

Abstract: Recently Affine Projection Algorithm（APA） has become a very popular algorithm in the application of acoustic echo cancellation. According to the contradictory factors of convergence rate and steady-state misalignment, a new algorithm by the behavior of associating variable regularization and evolving order has been proposed. It will reach a much better state by coordinating variable regularization and evolving order. Instead of the conventional assumption that the a posteriori error is zero, this paper incorporates the statistical characteristic of the noise into the adaptation process. Exact and approximate formulations for the optimal regularization factor are derived. Numerical simulation results show that the proposed algorithm improves the performance of the APA in terms of its faster convergence rate and lower steady-state misalignment compared to existing variable regularization APA and evolving order APA, respectively.

Key words: variable regularization, evolving order, Affine Projection Algorithm（APA）, acoustic echo cancellation

冀常鹏1，2，姬红红1，郭伟平1. 变阶变正则化APA在回声抵消中的应用[J]. 计算机工程与应用, 2014, 50(15): 168-171.

JI Changpeng1，2, JI Honghong1, GUO Weiping1. APA with evolving order and variable regularization for echo cancellation[J]. Computer Engineering and Applications, 2014, 50(15): 168-171.