摘要
针对传统的频域批处理LMS(Frequency-domain Block Least Mean Square,FBLMS)算法在收敛速度和稳态误差之间存在矛盾的问题,不同于变步长LMS算法,提出了一种新的变块长频域批处理LMS算法,采用自适应改变的批处理块块长的方法来协调解决这个矛盾。通过Matlab对提出的算法进行计算机仿真,结果表明相比于传统的FBLMS算法,新算法具有更快的收敛速度和更小的稳态误差。
To conquer the confliction of convergence speed and steady stage error in the traditional FBLMS （Frequency-domain Block Least Mean Square） algorithms,a new variable block-size frequency-domain LMS adaptive algorithm is proposed,which is different from variable step-size LMS adaptive algorithms.It balances the conflict by adjusting block-size adaptively.The simulations on a Matlab platform show that the proposed algorithm has higher convergence speed and smaller steady stage error than the traditional FBLMS algorithms.
出处
《计算机工程与应用》
CSCD
北大核心
2009年第2期34-38,共5页
Computer Engineering and Applications
基金
国家高技术研究发展计划(863)No.2007AA01Z230~~
关键词
频域
最小均方(LMS)算法
变块长
自适应滤波
frequency-domain
Least Mean Square（LMS） algorithm
variable block-size
adaptive filtering
