Combined Regularization Affine Projection Sign Algorithm Against Impulsive Noises

The affine projection sign algorithm with the regularization parameter has a trade-off problem between the fast convergence and the small steady-state error. In this paper, we propose a combined regularization affine projection sign algorithm to resolve this trade-off. The variable mixing factor is composed of sigmoidal activation function and used to mix two affine projection sign algorithm with different regularization parameters. To make the algorithm be robust to the impulsive noise, the mixing factor is updated indirectly by minimizing the L-1 norm of a priori error. Simulation results show that the proposed algorithm shows better performance than the conventional algorithms.