Robust adaptive beamforming based on reconstruction of interference covariance matrix and estimation of steering vector

An innovative beamforming strategy, which integrates the reconstruction of the interference-plus-noise covariance matrix (INCM) with the estimation of the steering vector (SV), is proposed in this study. By employing Capon spectral search techniques, the algorithm preliminarily identifies the directions of the signal of interest (SOI) and interfering signals. Considering the uncertainty of array calibration, an annular uncertainty set has been used in this paper to constrain the interference signal's nominal and actual errors of the SV, and Capon spectral integration is performed over this high-dimensional annular surface area. Given the complexity of integral calculations in high-dimensional areas, a discrete summation method has been adopted in this study to approximate the solution of the integral, thereby obtaining the INCM without the SOI. Furthermore, by utilizing subspace characteristics, vertical error vectors are constructed and a neighborhood table of SV errors for the SOI is created in this paper. This table is then used further to optimize the precise estimation of the SOI's SV. Ultimately, the weight vector for robust adaptive beamforming (RAB) is calculated using the corrected SV and the reconstructed INCM. The main advantage of this method lies in its robustness against various unknown mismatches, and it avoids solving the quadratically constrained quadratic programming (QCQP) problem, significantly reducing the complexity and computational demands of the algorithm. Simulation results validate the effectiveness and robustness of the strategy.