Research On Robust Adaptive Beamforming Based On Covariance Matrix Reconstruction

As the application range of array signal processing becomes more and more extensive,beamforming,as an important content in array signal processing,has received more and more attention due to the rapid development of wireless communication.In practical applications,the working environment is much more complicated than expected,and there are various errors,such as array calibration errors,signal waveform distortion,and direction of arrival errors.Besides the number of snapshots is usually limited.These factors make the performance of traditional adaptive beamforming algorithms severe decline,so it is necessary to improve the robustness of the beamforming algorithm.This paper studies and does the following work in depth:1.The background and significance of beamforming algorithms are deeply studied.We establish an array receiving signal model under ideal conditions and expound the commonly used criteria in adaptive beamforming algorithms.The sources of errors in adaptive beamforming algorithms and the reasons of performance degradation are analyzed.2.Aiming at the performance degradation of the traditional adaptive beamformers when the sampled covariance matrix contains the expected signal and the desired signal steering vector mismatches,an improved robust beamforming algorithm based on covariance matrix reconstruction is proposed.The algorithm first estimates the spatial spectrum and integrates the approximate direction range of the desired signal and interference to estimate the steering vector.Then dominant mode rejection is used to remove the redundant correlation between the signals.Finally the interference-plus-noise covariance matrix is reconstructed.The algorithm effectively improves the robustness when the desired signal steering vector mismatches,and has lower complexity.3.Aiming at the performance degradation of the traditional adaptive beamformers due to the difference between the sampling covariance matrix and its statistical covariance matrix at low snapshots,we considered an extreme situation of single snapshot and proposed a single snapshot robust adaptive beamforming algorithm based on compressed sensing.The algorithm first uses the sparseness of the array receiving signal to reconstructs the sparse signal by the OMP algorithm in compressed sensing.Then the steering vector of the desired signal and interference can be estimated to reconstruct the full-rank interference-plus-noise covariance matrix so that the covariance matrix is invertible.The algorithm effectively improves the robustness of the algorithm,reduces the sensitivity to snapshots,and has lower complexity.4.Aiming at the performance degradation of the traditional adaptive beamformers due to the difference between the sampling covariance matrix and its statistical covariance matrix at low snapshots,another single snapshot robust beamforming algorithm based on SPICE is proposed.The algorithm firstly uses the SPICE algorithm to iteratively estimate the parameters needed to reconstruct the interference covariance matrix,and then estimates the noise power by solving a linear programming problem to reconstruct a more accurate full-rank interference –plus-noise covariance matrix.Finally the steering vector of the desired signal is modified.The algorithm effectively improves the robustness of the algorithm and reduces the sensitivity to the number of snapshots.