Simplify adaptive IIR notch filter based on least mean p-power error criterion

Abstract

In this paper, a simplify second order IIR adaptive notch (ANF) algorithm based on least mean p-power (LMP) error criterion is presented. The quantizing gradient technique is used to detect high convergence speed and high impulsive noise robustness. This algorithm reduces an amount of multiplication computational equal to the length of input data. It is proved here that the convergence speed, variance, bias and noise robustness do not depend on p values. For practical application with most desirable purposes, the value of p=1 is chosen, the results of computer simulation and application on TMS320C50 digital signal processing starter kit are given to demonstrate the convergence speed and noise robustness performance.