Robust Adaptive Beamforming via Generalized Linear Combination of Covariance Matrix with Eigenvalue Decomposition Refinement

The article delves into the critical challenges faced by traditional robust adaptive beamforming (RAB) methods, particularly the difficulty in removing the cross-component between the desired and interference signals. It introduces an innovative approach that leverages the general linear combination (GLC) method for autonomous diagonal loading, significantly enhancing the reconstruction of the interference-plus-noise covariance matrix (INCM). The proposed method employs a block matrix constructed using conventional beamforming to suppress the desired component, followed by eigenvalue decomposition to further refine the INCM reconstruction. This dual-stage process ensures a more accurate and robust beamforming solution. The article also presents extensive simulation results that highlight the superior performance of the proposed method under various mismatch conditions, including coherent and incoherent errors, amplitude and phase errors, and steering vector random errors. These results demonstrate the method's effectiveness in enhancing the signal-to-interference-plus-noise ratio (SINR) and its robustness against different types of mismatches, making it a significant contribution to the field of adaptive beamforming.