A comparison between adaptive anc algorithms with and without cancellation path modelling

The adaptive filters in active noise control (ANC) systems differ from other common adaptive filters in the existence of the cancellation path, which is the transfer function between the outputs of the adaptive control filters and the error sensors. Cancellation paths play a critical role in active noise control systems, and the corresponding adaptive algorithms usually require the information of the cancellation paths for updating the control filters. The most commonly used filtered-x LMS algorithm takes into account the cancellation paths by filtering the reference signal with an estimate of the cancellation path transfer functions. For many ANC applications, the cancellation path modelling must be carried out online to maintain the stability of the system, and one modelling method obtains the cancellation path information by injecting uncorrelated signal into the cancellation path. This paper will introduce the filtered- x LMS algorithm embedded with this online cancellation path modelling and the direction search LMS algorithm, which is one of the ANC algorithms that do not need an explicit model of the cancellation path. In the direction search LMS algorithm, the standard LMS algorithm is adopted to update the adaptive filter coefficients directly with the reference signal by automatically choosing a proper update direction based on the monitoring of the excess noise power. The performance of the two typical adaptive ANC algorithms, one with and one without cancellation path modelling, will be compared in terms of noise reduction level, tracking speed, computation load and robustness.