Factorized All-Pass Based IIR Adaptive Notch Filters

This paper introduces a new family of infinite impulse response adaptive notch filters that forms multiple notches using a second-order factorization of an all-pass transfer function. The new orthogonal realization is amenable for adaptive filtering to obtain the unknown frequencies of interest. Two new adaptive filtering algorithms are presented that can achieve fast convergence at low computational cost. Local convergence analysis for the new algorithms is performed, and a detailed discussion of their properties is provided. The new all-pass based notch realization introduces a different compromise between bias and signal-to-noise ratio (SNR) when compared with realizations previously reported in the literature. Specifically, it achieves lower bias than other approaches at low SNR. This property is particularly attractive for the estimation and tracking of multiple sinusoids. Furthermore, the bias can be made arbitrarily small or can be accurately estimated and compensated for. Extensive computer simulations are provided to illustrate the performance of the proposed adaptive notch filters in terms of bias, speed of convergence, and tracking capability.