基于Delta算子的SVD-LMS算法  

A SVD-LMS Algorithm Based on Delta Operator

作　　者：宋雪洁[1] 张端金[1] 王香丽[1] 

机构地区：[1]郑州大学信息工程学院,河南郑州450052

出　　处：《现代电子技术》2006年第11期79-80,83,共3页Modern Electronics Technique

基　　金：河南省自然科学基金资助项目(0311011600);河南省高校青年骨干教师资助项目(2003100)

摘　　要：由于传统的最小均方算法在性能上存在着一些缺陷,比如收敛速度慢等,这都会影响到整个系统的性能,因此研究了一种基于Delta算子描述的改进自适应滤波算法。利用矩阵的SVD(奇异值)分解技术推导出了Delta LMS算法的迭代步骤,这种算法能够降低设计的滤波器阶数,有效地减少计算量,并提高了收敛速度,改善了对误差的跟踪性能。运用Matlab对Delta SVD LMS算法进行仿真并与Delta LMS算法进行了比较,结果表明该算法的有效性。A modified adaptive filter decomposition is studied based on Delta operator, because traditional LMS in abilities have some shortages,for instance its lower convergence,which affect the quality of whole systems. The recursive formulae of the Delta - SVD - LMS is developed, the algorithm can descend effectively the quantity of the filter and the computation. It has high convergence,and the stability is improved. Matlab used to the simulation, the results indicate Delta - SVD - LMS algo- rithm is effective.

关 键 词：DELTA算子 SVD分解 LMS自适应滤波 收敛速度 

分 类 号：TN911.7[电子电信—通信与信息系统]