改进的l_0范数LMS算法与分析

A Modified l_0_LMS Algorithm and Its Performance Analysis

提出一种改进的基于l0范数的最小均方(LMS)算法.采用误差的相关函数值调整权系数步长因子以及零吸引项,增强系统的抗噪声性能;并且引入一种修正的权系数步长因子更新方法,进而使系统具有较快的跟踪速度.对提出的算法进行理论分析,最后在不同信噪比下进行仿真验证并与已有的基于l0范数的LMS算法进行比较.理论分析结合仿真验证都表明新提出算法具有较快的跟踪速度和较强的抗噪声性能.

A new variable step-size l0_least mean square（ LMS） algorithm is proposed. A step size control method and the zero attraction items reweight method based on correlation function value of the error to increase the convergence speed,and reduce the steady-state misalignment. The anti-noise performance,convergence,tracking steady state error and misadjustment of this algorithm are discussed in theoretical analysis. Finally,the algorithm is compared with l0_LMS and Il0_LMS in different signal-to-noise ratio. Theoretical analysis combined with experimental simulation conclusion： the algorithm can achieve better tracking speed,lower steady state error and anti-noise performance.