摘要
在对传统LMS算法、变步长SVSLMS算法及归一化LMS算法分析的基础上,提出了一种改进的归一化变步长LMS算法即N-SVSLMS(Normalized-SVSLMS)算法。该算法结合了参考文献中两种算法的思想,得到了改进的归一化LMS自适应算法。该算法在信道环境多变的情况下,收敛速度和稳定性能有了进一步的提高。理论分析及计算机仿真结果表明,N-SVSLMS算法明显优于传统LMS算法、变步长SVSLMS算法及归一化的LMS算法。
This paper based on the analysis of common LMS algorithm,variable step-size LMS algorithm and normalized LMS algorithm,proposes an improved variable step-size LMS adaptive filter algorithm.The algorithm combines the ideas in literature and literature,thus achieving the improved normalized LMS adaptive algorithm.Theoretical analysis and computer simulations show that its convergence speed and performance stability is further improved in the changeable channel environment,and the proposed algorithm is remarkably better than the traditional LMS algorithm,variable step-size LMS algorithm and normalized LMS algorithm.
出处
《通信技术》
2010年第11期153-155,159,共4页
Communications Technology
基金
国家自然科学基金资助项目(批准号:70271001)
国家博士后基金资助项目(No.2002032191)
上海市科委重点计划资助项目(No.05dz05036)
关键词
变步长
归一化
自适应滤波
LMS
variable step-size
normalized
adaptive filter
LMS
