Improved subspace processing for robust adaptive beamforming based on covariance matrix reconstruction
Song Wang, Defu Jiang, Yan Han · 2025
Conventional adaptive beamformer impose stringent requirements on model accuracy. If the angle of the desired signal (DS) or interference-plus-noise covariance matrix (IPNCM) is inaccurately estimated, the output signal-to-interferenceplus-noise ratio (SINR) could be severely degraded. Thus, we propose an improved subspace processing algorithm to reconstruct IPNCM and DS steering vector (SV). The algorithm includes two steps. First, we employ subspace projection technique to project the estimated DS SV onto the DS subspace, so as to obtain a more accurate DS SV. Second, the interference signal SV and power are re-estimated using subspace techniques, then we re-construct the IPNCM by spectrum estimation theory, which either does not contain or contains fewer expected signal components. Next, the weighting coefficients of the array are obtained with the solution formula of the MVDR beamformer. Numerical simulation indicates that the algorithm can attain a near-optimal output SINR with smaller computational complexity compared to other algorithms.