Dominant Pole Localization of FxLMS Adaptation Process in Active Noise Control

This paper develops a technique for dominant pole localization of the adaptation process, performed by the FxLMS algorithm in active noise control systems. This development results in a new adaptation algorithm, called Filtered Weight FxLMS (FwFxLMS). Similar to the FxLMS, FwFxLMS uses a recursion for updating the weight vector of a transversal ANC controller. However, in the new algorithm, the weight vector is filtered before being updated by the recursion. This filtering procedure causes dynamic behaviors of the adaptation process to be improved. This improvement can be theoretically proved by plotting and analyzing the root locus of the adaptation process in the z-plane. Also, simulation results show the efficiency of the proposed algorithm in active control of acoustic noise.