A Variable Step-Size Affine Projection Algorithm With Periodically Evolved Update Interval

Abstract

The proposed is an improved affine projection algorithm based on two kind variables. The first variable is the step size. The change of that is appeared at the criterion on the projected weighted error vector. The second variable is the evolving period. The update interval of the weight vector is changed with the analysis of mean square deviation which detects the cross point of the transient stage and the steady-state stage. Simulation result shows that the proposed algorithm is improved with the convergence rate and the steady-stage error with the variable step-size and the update interval.evolving period. The update interval of the weight vector is changed with the analysis of mean square deviation which detects the cross point of the transient stage and the steady-state stage. Simulation result shows that the proposed algorithm is improved with the convergence rate and the steady-stage error with the variable step-size and the update interval.