卫星通信中一种改进的变步长LMS均衡算法研究  

A modified variable step size LMS adaptive filtering algorithm in satellite communication

作　　者：魏武 左天虎 刘期烈[2] 黄巍[2] 李云[2] 

机构地区：[1]国家移动卫星通信工程技术研究中心,南京210000 [2]重庆邮电大学移动通信技术重点实验室,重庆400065

出　　处：《重庆邮电大学学报（自然科学版）》2014年第1期42-48,共7页Journal of Chongqing University of Posts and Telecommunications(Natural Science Edition)

基　　金：国家"863"项目(2012AA01A507);国家重大专项(2013ZX03006003-002);国家自然科学基金(61071118;60702055)~~

摘　　要：针对传统的定步长最小均方误差(fixed step size-least mean square,FSS-LMS)算法不能同时兼顾快速收敛和较小稳态误差,以及变步长最小均方误差(variable step size-least mean square,VSS-LMS)算法也不能满足较快的收敛速度和较好的抗噪声性能等问题,提出了一种在Sigmoid函数中引入递减的等比序列,在估计误差的自相关函数基础上,用前后时刻的绝对误差来代替当前时刻的瞬时误差,加强了对步长因子控制的算法。仿真结果表明,在不同的信噪比环境下,提出的算法较其他LMS算法具有收敛速度快、抗噪声性能强和稳态误差小等特点。同时,给出了卫星通信系统的仿真模型,并且将提出的算法应用到了该系统模型的自适应均衡器之中,系统的误码性能有较大的改善。In order to solve the problem that the conventional fixed step size LMS （FSS-LMS） algorithm cannot consider both the convergence speed and steady state error simultaneously , and the variable step size LMS （VSS-LMS） algorithm cannot manage the conflict between fast convergence speed and anti-noise performance, too, we introduce a decreasing geo- metric sequence into the Sigmoid function and make full use of the absolute error of the current moment and the previous based on autocorrelation function of the estimation error instead of the current moment only to strength the control of the step size factor. The simulation results show that compared with several LMS algorithms, our algorithm has faster convergence speed, lower steady state error and better anti-noise performance. At the same time, this paper proposes the satellite com- munication transmission system simulation model, and shows us the great improvement of the system bit error rate （BER） performance when the new algorithm is applied to the adaptive equalizer of the system model.

关 键 词：卫星通信系统 自适应滤波 变步长 LMS算法 抗噪声性能 

分 类 号：TN927.2[电子电信—通信与信息系统]