Direct-Lattice Adaptive Notch Filter for Frequency Estimation and Tracking

Abstract

This paper presents a direct-lattice adaptive notch filter (DLANF) for frequency estimation and tracking. The proposed filter is the cascade of an all-pole direct form structure filter with an all-zero lattice predictor. The normalized gradient algorithm is used to adjust the filter parameter. This technique outperforms some conventional adaptive filters in terms of convergence time and provides a small mean square error (MSE). Computer simulations are conducted to show the superiority of the proposed filter.