A sparseness controlled proportionate algorithm for acoustic echo cancellation

Sparseness variation in acoustic impulse response arises due to changes in temperature, pressure, acoustic source movements and changes in the acoustic environment. Therefore, the algorithms employed in acoustic echo cancellation have to perform well for both sparse and dispersive unknown systems. The well-known algorithms, normalised least mean square (NLMS) or proportionate NLMS (PNLMS), are limited to perform well either in dispersive or sparse cases, respectively. The proposed sparseness-controlled PNLMS (SC-PNLMS) algorithm inherits the beneficial properties of both PNLMS and NLMS by employing the sparseness measure into the PNLMS algorithm. Simulation results presented show improved performance over the PNLMS algorithm even for dispersive impulse responses. copyright by EURASIP.