Abstract
The filtered-x least mean square (FxLMS) algorithm for active noise control (ANC) systems requires filtering of the reference signal via the cancellation path modeling (CPM) filter. The CPM filter may be available from the offline experiments. However, in many practical situations the online estimation of CPM filter may be needed due to the possible time-varying nature of the cancellation path. This paper investigates estimation of cancellation path during the online operation of the FxLMS-based feedback-type ANC system. The key idea in the proposed approach is to employ a gain-controlled probe noise mixed with the cancelation signal. The gain-factor for the probe noise is adjusted such that a large-level probe signal is used during the transient state of the ANC system. This improves the convergence of the cancellation path estimation process. As the ANC system converges, the level of the probe noise is reduced to achieve a good noise reduction performance. Simulation results demonstrate the effectiveness of the proposed method.