Robust adaptive beamforming based on interference subspace projection in GSC

Abstract

In this thesis, we propose a new robust adaptive beamforming algorithm applicable to an on-frequency repeater system with antenna array. In the on-frequency repeater system, the back-lobe interference cancellation and system performance degradation caused by array error are the most important problems. The proposed beamforming algorithm makes generalized sidelobe canceller be robust. This method is to project the weight vector into interference subspace in order to solve the above problems. This interference projection matrix prevents the desired signal cancellation caused by array error. It can be obtained from the eigenvectors of the estimated covariance matrix constructed from the blocking matrix output and is updated whenever the covariance matrix and weight vector is updated. We employ singular-value decomposition and the row vectors of covariance matrix as the method to find eigenvectors, and then we compare the performance each other. Also, we propose the scheme of the proposed adaptive beamforming algorithm with expanded array output, and compare the performance of it with it of proposed adaptive beamforming algorithm with array output only.