LMS自适应滤波算法的改进及性能分析

LMS Adaptive Filtering Algorithm Improvement and Performance Analysis

关于研究自适应滤波器算法优化问题,为了改善数字传输系统抗干扰能力,提高稳定性,减小误差,引入了自适应滤波算法,并详细分析了算法步长因子与失调误差e(n)之间关系对算法性能的影响。采用了改进的双曲正切函数为变步长,对传统的LMS算法进行了改进。分析了参数α、β、m对算法性能的影响。算法不但收敛速度快、稳态误差小,而且克服了变步长LMS算法达稳态时,步长变化太快的缺陷。利用MATLAB仿真结果与理论分析相一致,表明了算法具有良好的性能特性,可广泛应用于各种数字通信中。

In order to improve anti-jamming capability of the digital transmission system,this paper introduced an adaptive filtering algorithm and analyzed the effect of the relation between step-size factor and e（n） on algorithm performances.The traditional LMS algorithm was improved,which used a modified hyperbolic tangent function for the variable step-size.In this paper,how the parameters of,β,and m affected the performance of the novel algorithm was discussed.The algorithm can not only obtain fast convergence and lower steady-state error,but also avoid the shortage of fast step change in the process of the adaptive steady state.MATLAB simulation results confirmed the theoretical analysis,which demonstrated that the algorithm has better performance and can be widely used in various digital communication.