基于向量图分析的自适应滤波快速算法

Fast algorithms of adaptive filtering based on vector plots analysis

对自适应滤波算法进行了讨论,提出了基于向量图分析的快速算法.该方法与目前所有自适应滤波算法不同,将数学中的几何分析方法引入到自适应滤波的研究中,通过探讨最小均方(LMS)算法的向量图结构及其算法收敛的几何特征,在基于几何分析的基础上,寻找有效的快速算法.仿真结果表明了所获算法的有效性及优越性,从而为自适应滤波算法的研究开辟了另一条新的途径.

Adaptive filtering algorithms were discussed and a fast algorithm was proposed based on vector plots analysis, which was different from the present adaptive filtering algorithms. It applied the approaches of mathematical geometrical analysis into the study of adaptive filtering and tried to seek an effective and fast algorithm on the basis of geometrical analysis, through inquiring into vector plots structure of least mean squares (LMS) algorithm and geometrical feature of algorithm convergence. Numerical simulations indicate the efficiency and superiority of this new algorithm. Thus, a new approach is in prospect for the research of adaptive filtering algorithms.