一种变步长LMS自适应滤波的改进算法  

An Improved LMS Adaptive Filtering Algorithm with Variable Step Size

作　　者：吴瑶 张海霞 WU Yao;ZHANG Haixia(Unit 91033 of PLA,Qingdao Shandong 266001,China)

机构地区：[1]解放军91033部队,山东青岛266001

出　　处：《通信技术》2021年第2期307-311,共5页Communications Technology

摘　　要：在信号处理领域,变步长LMS自适应滤波算法被广泛应用,其主要优点为具有较强的鲁棒性和简单易实现。但是,该算法在收敛速度上总是差强人意。针对这个问题对现有的变步长LMS自适应滤波算法进行了改进。首先,介绍了典型LMS算法和传统变步长LMS自适应滤波算法的基本原理及其局限性,在此基础上针对现有基于Sigmoid函数变步长最小均方算法,引入一个调节函数以进一步增强算法的稳定性。仿真实验结果表明,改进算法在保持传统算法精度的情况下,较大程度上提高了算法在大误差情况下的收敛速度,具有较强的实用价值。In the field of signal processing, the variable step LMS adaptive filtering algorithm is widelyused. Its main advantages are strong robustness and easy implementation. However, the algorithm isalways unsatisfactory in terms of convergence speed. In order to solve this problem, the existing variablestep LMS adaptive filtering algorithm is improved. Firstly, the basic principles and limitations of typicalLMS algorithms and traditional variable-step LMS adaptive filtering algorithms are introduced. Then, onthis basis, for the existing Sigmoid function-based variable step LMS algorithm, an adjustment function isintroduced to further enhances the stability of the algorithm. The simulation experiment results indicate thatthe improved algorithm greatly improves the convergence speed of the algorithm in the case of large errorswhile maintaining the accuracy of the traditional algorithm, and has strong practical value.

关 键 词：自适应滤波 变步长LMS算法 收敛性 稳态性 调节函数 

分 类 号：TN911.7[电子电信—通信与信息系统]