A variable step-size proportionate affine projection algorithm for echo cancellation

The sparseness-controlled improved proportionate affine projection algorithm (SC-IPAPA) has been proposed recently to accelerate convergence rate for acoustic both network and echo cancellation in both sparse and dispersive circumstances. Unfortunately, there are conflicting objectives between fast convergence and low mis-adjustment for the algorithm with fixed step size parameter. In this paper we proposed a variable step-size version of it, namely VSS-SC-IPAPA, providing a feasible solution for this problem. The new algorithm obtained a new step size control approach for proportionate affine projection algorithm by forcing the a posteriori error not to be zero, but to be the disturbance signal. It incorporates the advantages of the sparseness-controlled algorithm, data reusing method and variable step size algorithm. Echo cancellation simulation results confirm that the proposed algorithm can constitute a significant improvement in the convergence speed with very small mis-adjustment when compared with SC-IPAPA employing fixed step-size constant parameter. Furthermore, this algorithm offers the overall best performance compare to other variable step-size algorithms across the range of sparseness measure.