基于改进的双曲正切函数变步长LMS算法

为了改进现有的变步长最小均方误差(least mean square, LMS)算法在低信噪比时性能较差的缺陷，提出了一种基于改进的双曲正切函数的变步长LMS算法，从理论分析和仿真实验两方面讨论了引入参数对算法收敛性、跟踪性、稳定性的影响及算法的抗干扰性。理论分析和仿真实验表明该算法在高低信噪比时均具有较快的收敛速度和跟踪速度以及较小的稳态误差和稳态失调，并且在低信噪比时该算法的收敛性、跟踪性、稳态性均优于其他多种变步长算法。

New variable step-size LMS algorithm based on modified hyperbolic tangent function

In order to improve the overall performance of variable step size least mean square (LMS) algorithms under the low signal-to-noise ratio (SNR) condition, a new variable step size LMS algorithm based on modified hyperbolic tangent function is presented. The anti-noise performance of this algorithm and the influence exerted by the introducing parameters on convergence, tracking, steady state error and misadjustment are discussed minutely in both theoretical analysis and simulation experiment. The theoretical analysis and simulation results show that this algorithm can achieve both better convergence speed, tracking speed，lower steady state error and misadjustment under high and low SNR conditions. The results also show the overall performance of this algorithm exceeds some others under low SNR condition.