A Novel Robust Adaptive Beamforming Algorithm Based on Subspace Orthogonality and Projection
Jiayu Guo, Huichao Yang, Zhongfu Ye
Abstract
Recently, it has been extensively researched about designing robust adaptive beamforming (RAB) algorithms to deal with model mismatch issues. In this article, a RAB algorithm is proposed to estimate the steering vectors (SVs) of the incident sources and reconstruct the interference-plus-noise covariance matrix (INCM). First, we construct the error SVs in the noise subspace, which correct the nominal SVs by iterative updates to obtain a more accurate estimation of SVs. Then the projection matrix is constructed utilizing the estimated SV of the signal of interest (SOI), and the interference powers are estimated by projecting the sampled covariance matrix (SCM). Furthermore, two virtual interferences are added on both sides of each estimated interference direction to widen the nulls in the corresponding directions of the interferences. Finally, the INCM can be constructed and utilize the estimated SV of the SOI to calculate the weight vector of the beamformer. The proposed method has less computational complexity and the simulation results show that the proposed method is more robust to various types of mismatches in comparison to previous algorithms.