摘要： 在研究归一化最小均方误差(NLMS)算法的基础上，提出一种基于指数函数的变步长LMS算法。通过建立误差 和步长 的函数关系，实时调整步长，并对输入信号完成时域信号解相关，解决稳态失调系数与收敛速度的矛盾。仿真实验结果证明，该算法与传统LMS算法、SVS_LMS算法、NLMS算法以及双曲正切变步长LMS算法相比，具有更高的收敛速度和较小的稳态失调系数。

关键词: 指数函数, 归一化, LMS算法, 变步长, 解相关, 稳态失调

Abstract: This paper proposes a new variable step size Normalized Least Mean Square(NLMS) algorithm based on exponential function according to NLMS. A function relationship is established between signal error and step size μ, and the contradiction between convergence speed and maladjustment error is solved. Simulation experimental results show that the ameliorative LMS algorithm has faster convergence speed and smaller maladjustment error than the commonly LMS algorithm and a variable step size LMS algorithm based on hyperbolic tangent function.

Key words: exponential function, normalization, Least Mean Square(LMS) algorithm, variable step size, decorrelation, steady-state misadjustment

中图分类号: 

• TP18