文件名称:LMS
介绍说明－－下载内容均来自于网络，请自行研究使用
在一种变步长算法基础上, 从语音信号相关性的角度出发, 提出了一种新的去相关变步长LMS 算法( DCL—NLMS) 。该算
法结构简单, 收敛速度快, 稳态失调小, 计算量与NLMS 算法相当。仿真结果表明, 该算法在处理强相关性信号时, 不仅收敛速度明
显快于其余算法, 而且稳态失调特性也有很大优势-Based on the variable step NLMS arithmetic, this paper proposes a new uncorrelated variable step arithmetic( DCL—
NLMS) by employing auto—correlation of the input speech signal.The proposed algorithm has simple structure, fast convergence
speed, but small stability maladjustment with the same computational complexity as NLMS.Simulation results show not only its
convergence is faster than that of the others, but also the stable maladjustment of it is better too.
法结构简单, 收敛速度快, 稳态失调小, 计算量与NLMS 算法相当。仿真结果表明, 该算法在处理强相关性信号时, 不仅收敛速度明
显快于其余算法, 而且稳态失调特性也有很大优势-Based on the variable step NLMS arithmetic, this paper proposes a new uncorrelated variable step arithmetic( DCL—
NLMS) by employing auto—correlation of the input speech signal.The proposed algorithm has simple structure, fast convergence
speed, but small stability maladjustment with the same computational complexity as NLMS.Simulation results show not only its
convergence is faster than that of the others, but also the stable maladjustment of it is better too.
(系统自动生成,下载前可以参看下载内容)
下载文件列表
一种新的去相关变步长LMS算法及其应用.pdf