A variable parameter improved proportionate normalized LMS algorithm

A proportionate adaptive algorithm is proposed to improve the convergence of the improved proportionate NLMS (IPNLMS) algorithm for various kinds of impulse response. Although we know that the convergence speed of a proportionate algorithm depends on the sparsity of the target impulse response, there is no clear knowledge on their relationship. A real function is firstly established here to characterize the relationship between the sparsity of the impulse response and the parameter of IPNLMS and then a variable parameter IPNLMS algorithm is proposed. The proposed algorithm detects the sparsity of the adaptive filter, and then adjusts the parameter of IPNLMS according to the sparsity. Simulation results show that the proposed algorithm can improve the convergence speed of IPNLMS in any case of sparsity.