Abstract:

Focusing on the performance reduction problem of the adaptive beamformer when the desired signal (DS) steering vector mismatch exists, an interference-plus-noise covariance matrix (INCM) reconstruction method via DS robustly blocking is proposed. First of all, the signal blocking matrix with angle broadening is constructed to separate the DS from the echo data, thus the quasi INCM is calculated with the uncontaminated echo data. Then after applying Eigen-decomposition on the quasi INCM, the INCM is accurately reconstructed by means of matrix projection transition. Finally, the effective estimation of the DS steering vector is completed by the way of eigen-decomposing the sample covariance matrix, of which the interference-plus-noise component has been eliminated. Theory analysis and simulation result show that the proposed algorithm has not only almost ideal performance but also low complexity.

Key words: beamforming, covariance matrix reconstruction, desired signal (DS) blocking, low complexity