Array steering vector estimation with nonstationary noise exploiting calibrated noise subspace

Abstract

Most of the robust adaptive beamformers to resolve target signal cancellation caused by array steering vector (ASV) mismatch use prior information on ASV. However, in some practical applications, no prior information on ASV is available or the ASV is changing due to the movement of target source. In such case, the identification of ASV is essential. There are some methods proposed for ASV identification but few of them work with nonstationary noise. In this paper, a new method is proposed to identify ASV with nonstationary noise by exploiting the calibrated noise subspace. With the identified ASV, a robust beamformer can be formed. Experiments illustrate that the proposed method has high performance in adverse environment.