摘要
在无线直放站反馈干扰抵消的过程中,自适应滤波器的误差信号可以接收目标信号与残余回波的混合,是阻碍滤波器根据残余回波强度,快速调整抽头系数.利用误差信号的特点,给出了一种基于信噪比的改进变步长最小均平方(least mean square,LMS)自适应算法.该算法利用误差信号和滤波器的输出信号作为目标信号和反馈干扰信号的估计,根据目标与干扰信号的功率比值来调整自适应滤波器的步长.计算机仿真结果表明,该算法具有快速的初始收敛速率和较小的超量均方误差.在稳态情况下,对于干扰的突变能够快速地再次收敛,表明该算法在反馈干扰抵消方面的性能优于已有的算法.
During the feedback interference cancelling of wireless repeaters, the error signals of adaptive filters are mixtures of target signals and residual echoes,which hinder the filters from rapidly adjusting their coefficients according to the strength of residual echoes. An improved SNR（signal noise ratio） based variable step size LMS（least mean square） adaptive algorithm is presented using the characters of the error signals. The algorithm makes use of the error signals and the filters＇ output signals as the estimation of target signals and feedback interference signals,and adjusts the adaptive filters＇ coefficients according to the power ratio of target and echo signals. Computer simulations demonstrate that the proposed algorithm has a rapid initial convergence speed and a small excess mean square error. The algorithm is able to converge again rapidly in the case of an abrupt change of interference when it has been in a stable state,which confirms that the proposed one is superior to former algorithms in terms of the cancelling performance.
出处
《厦门大学学报（自然科学版）》
CAS
CSCD
北大核心
2011年第5期829-833,共5页
Journal of Xiamen University：Natural Science
基金
福建省科技重大专项项目(2009HZ0003-1)
福建省教育厅科技项目(JA10251)
(JB10146)
厦门市科技计划项目(3502Z20103038)
关键词
反馈干扰抵消
变步长
最小均平方算法
feedback interference cancelling
variable step size
LMS（least mean square） algorithm
