改进变步长LMS算法及在自适应噪声抵消中的应用

在分析传统LMS算法、变步长LMS算法及其改进算法的基础上,通过建立步长因子μ与误差信号е之间的非线性关系,提出了一种改进的变步长LMS自适应滤波算法,并将其应用于自适应噪声抵消中,还分析了参数对算法性能的影响,结果表明该算法具有较快的收敛速度和较小的稳态失调。文中最后给出了仿真结果,仿真结果与理论分析一致。

Improved LMS Algorithm with Variable Step-Size and Its Application in Adaptive Noise Cancellation

Based on the analysis of the traditional LMS and the LMS algorithm with variable step-size,an improved adaptive LMS filtering algorithm with variable step-size is presented by building a nonlinear function of the relationship between step-size factor μ and error signal e.Then this method is applied to adaptive noise cancellation.It is found that the algorithm has a faster convergence speed and less stability maladjustment by analyzing the impact of the parameters on algorithm performance.Finally,simulation results are provided and compared with those of theoretical analysis.A good agreement between them is found.