Improved Adaptive Convex Combination of Least Mean Square (LMS) Algorithm

Abstract

In the normal adaptive convex combination of least mean square algorithm (CLMS), the rule for modifying mixing parameter is based on the steepest descent method. When the algorithm converges, it will generate zigzag phenomena, which can make the convergence speed become slowly. To solve this problem, a new method that combines steepest descent method with damp Newton method for the mixing parameter is presented in this paper. The improved method can get faster convergence speed as well as retain the properties of normal convex combination algorithm. The results of comparison and simulation verify that the improved method has faster convergence speed and better performance.