双变因子LMS自适应滤波算法

摘要

为提高最小均方(LMS)自适应滤波算法的性能,在步长因子与误差信号满足一种改进的Sigmoid函数关系式的基础上,提出了一种双变因子控制下的变步长LMS自适应滤波算法.该算法具有初始状态收敛速度快、稳定状态均方误差小、抗噪声性能好、适应系统跃变能力强的优点.计算机仿真结果验证了该算法的优越性.%A varariable setp-size LMS algorithm based on double variable factors is presented to improve the LMS (Least Mean Square) adaptive filtering algorithm performance. The improved Sigmoid functional relationship between the step-size and the error signal is established. The proposed algorithm has more faster convergence rate at the beginning and less steady-state MSE than the former algorithms. The adaptability of tracking system and deducing the effects of the irrelevant noise is another advantage of the improved algorithm. Computer simulation results confirm the algorithm is superior to the former algorithms in performance.