Variable step-size LMS adaptive algorithm based on computational verb rules

Abstract

In this paper, a new variable step-size LMS algorithm based on computational verb rules is presented. This method allows the algorithm to converge faster and to produce smaller steady-state errors compared to other variable step-size LMS algorithms, even when the system is subjected to sudden changes. Furthermore, it provides a brand new framework that allows the designers to design adaptive algorithms in a simpler and more effective way by using computational verb rules to encode the designers' knowledge of dynamical processes. Simulation results are provided to show the usefulness of the proposed algorithm by comparing with other variable step-size LMS algorithms.