摘要
对稀疏路径的自适应声回波对消算法进行了讨论,提出一种稀疏置零归一化解相关LMS自适应滤波算法(简称SSKNDLMS).该算法综合了稀疏算法,置零算法及归一化解相关算法的优点,在增加少量计算量的情况下,其收敛速度和稳态失配比常用的LMS算法都有明显的改善.计算机仿真证实了这一结果.
A hybrid sparse kill-weights adaptive filtering algorithm is proposed. This algorithm incorporates the advantages of the sparse algorithm, kill-weights algorithm and NDLMS algorithm. As a result, the improved algorithm demonstrates, at a little cost of computation load, a better performance both in convergence rate and misalignment than normal LMS algorithm in acoustic echo cancellation. The computer simulation resuits confirm the effectiveness of this algorithm.
出处
《四川大学学报(自然科学版)》
CAS
CSCD
北大核心
2006年第2期330-333,共4页
Journal of Sichuan University(Natural Science Edition)
基金
国家自然科学基金(60472096)
关键词
稀疏
置零
解相关
自适应声回波对消
sparse
kill-weights
decorrelating
adaptive acoustic echo cancellation