Variable step-size LMS algorithm: new developments and experiments

Abstract

The variable step size least mean square (VSLMS) algorithm is explored further and adopted for tracking of time varying environments. Two implementations of the VSLMS algorithm are proposed. Our emphasis is on the implementations with different step sizes at various taps of the adaptive filter. For an application where the adaptive filter is used to track a time varying plant, it is shown that the step sizes of the VSLMS algorithm converge towards their optimum values. Simulation results that support the developed theories are also presented.