New gradient-based algorithms for adaptive IIR notch filters

Abstract

This paper proposes two novel gradient-based algorithms for second-order adaptive IIR notch filters. They are based on a least mean p-power error criterion and a memoryless nonlinear gradient function. They are very attractive due to their computational efficiencies and improved performances. It is revealed by extensive simulations that they can produce significantly improved frequency estimates in both Gaussian and impulsive symmetric /spl alpha/-stable (S/spl alpha/S) noise scenarios compared with the existing gradient-type algorithms. Several simulated results are provided to show the superiority of the new algorithms.