<p>In this paper, a cross-weighted covariance matrix reconstruction (CWCMR) method is proposed to address the performance degradation of beamformers when the signal of interest (SOI) overlaps with interference signals. Since beamformers are designed to enhance the SOI while suppressing interference, weight vectors, specifically those from conventional beamforming and sample matrix inverse (SMI) beamforming, are applied within the interference region, which is efficiently identified using a Capon search. Based on this, a cross-weighted covariance matrix is constructed to suppress both the SOI and noise components. Furthermore, eigenvalue decomposition is employed to reconstruct a more accurate covariance matrix. Simulation results demonstrate that the proposed method outperforms several existing robust adaptive beamforming techniques under various mismatch conditions.</p>