Adaptive interference cancellation using microphone arrays

In this paper, an Interference Cancellation Algorithm using superdirective beamforming and adaptive noise cancelling (ANC) is proposed. The only assumption is that the source and the interference directions are coarsely known in advance. The noisy source signal of which the interference is reduced to some degree after beamforming, is processed by ANC in order to cancel out the residual interference left on the desired signal. In the proposed algorithm, the required noise-only reference signal for ANC is acquired by beamforming in the direction of the noise source. In this sense, proposed scheme is similar to well-known Generalized Sidelobe Canceller (GSC) algorithm. However, the directions of source and interference not need to be known exactly and it has much less implementation complexity compared to GSC. Thus, the proposed scheme is possible to be employed in microphone array systems with high number of microphones in real-time and without any need to know the directions, it can work in real life scenarios where knowing the exact positions are not realistic. The proposed scheme is verified by computer simutaions and real-life experiements.