Research On Improved Algorithm Of Adaptive Beamforming Based On Matrix Reconstruction

As an important branch of array signal processing,beamforming technology which gets the weight vector of the beamformer based on a certain criteria aims to receive the desired signals and suppress the interferences and noise.With the increasingly complex application scenarios of the beamforming technologies,the traditional beamforming technologies have been difficult to meet the needs of the actual harsh environment,and the robustness of the beamformer is seriously challenged.Therefore,the research on robust adaptive beamforming(RAB)algorithms for complex non-ideal environments has a very important theoretical and practical significance.In this dissertation,further research has been conducted on the basis of existing algorithms for covariance matrix reconstruction,and some effective algorithms have been proposed.The main contributions of this dissertation are briefly summarized as follows:1.Considering the influence of noise components on the reconstruction of covariance matrix,a robust adaptive beamforming algorithm is proposed to reconstruct the interference-plus-noise covariance matrix accurately.Firstly,the noise power can be estimated by Capon spatial spectrum in the angular region which only contains stationary noise,and the noise covariance matrix can be obtained.Then,a more accurate interference covariance matrix can be reconstructed by utilizing the Capon spatial power spectrum to collect all the interference information in the possible angular region of the interferences after removing the noise components in the integral region.Finally,using the reconstructed interference-plus-noise covariance matrix to estimate the SV of the desired signal,and a more robust adaptive beamfonner can be obtained.Compared with the existing reconstruction algorithms,the performance of the proposed algorithm has a huge improvement under various error conditions.2.Considering the errors existing in the process of covariance matrix reconstruction and SV estimation in the complicated non-ideal environment,a double-robust adaptive beamforming algorithm based on interference-plus-noise covariance matrix reconstruction and nonn-bounded constraints is proposed.The proposed algorithm utilizes that the errors of the reconstruction and estimation are norm-bounded.And then a further improvement has been made on the basis of the traditional Capon beamforming optimization problem.Finally,a more robust beamformer is obtained by solving a minimum-maximum optimization problem.3.A RAB algorithm based on covariance matrix reconstruction apply to low signal-to-noise ratio(SNR)conditions is proposed.First,we start from the covariance matrix of the desired signal and use the integral of Capon's spatial spectrum in the possible angular region of the desired signal as the reconstructed signal covariance matrix.Then,we can get the interference-plus-noise covariance matrix by subtracting the signal component from the sample covariance matrix combining with the least mean square error criterion.The robust adaptive beamforming algorithm not only narrows the integral region under multi-interference conditions and reduces the computational complexity,but also can well perceive the faint variation of the desired signal and eliminate the desired signal component from the sample covariance matrix while maintaining the robustness of the beamformer at low SNR.