Robust Adaptive Beamforming Based On Matrix Reconstruction

Adaptive Beamforming(ABF)uses prior information to adaptively adjust the weight of the antenna array to enhance the desired signal while suppressing interference and noise.In the Internet of Things era,the wireless communication scenarios are not noly exist in ordinary scenarios,but also involved in many complex and changeable scenarios,in which there are signal distortion,array deformation,phase fluctuation,and angle of arrival mismatch,etc.These phenomena will lead to inaccurate prior information and serverely affect the beamforming performance.Therefore,in practical applications,it is necessary to ensure that the ABF are robust enough to work in the presence of errors.Today’s robust adaptive beamforming algorithms have their own limitations.This thesis studies from the perspective of robustness to ensure the performance of the algorithm under non-ideal conditions.The specific work is as follows:In this thesis,the robustness of the ABF algorithm is studied from two aspects of correcting the covariance matrix error and broadening the beam null.Aiming at the defect that robust adaptive beamforming algorithm based on interference and noise matrix reconstruction(INCM)is fixed,a robust adaptive beamforming algorithm based on interference and noise matrix reconstruction in dynamic integral region(INCM-DIR)is proposed.The main idea of the algorithm is to change the originally selected fixed integration region to a dynamic integration region,so that even when the signal completely deviates from the integration region,an accurate covariance matrix can be reconstructed well;then the beam is broadened by constructing a projection subspace.The null width at the interference signal has an excellent tolerance for the error of the angle of arrival.The algorithm starts from the initially selected initial integration region,and judges whether there is a signal that can reconstruct the covariance matrix in the current region according to the power threshold calculated by the power spectral density.Appropriate integration regions,re-estimate the interference-plus-noise covariance matrix and the steering vector of the desired signal.The resulting reconstructed covariance matrix is used to widen the nulls.The simulation results show that in the presence of the angle of arrival error,the performance of INCM drops by more than 60%compared with the ideal situation,and the greater the angular deviation,the more severe the performance degrade.However,INCM-DIR has almost no obvious performance degradation whether there is a deviation in the direction of arrival or the number of snapshots is limited,and also create a widened null at the angle of the interference signal,and the robustness is significantly improved.The robust beamforming algorithm based on coprime array breaks through the limitations of the traditional uniform array in degrees of freedom and can receive large-scale signal sources,but the algorithm is not perfect in terms of robustness,and its performance will be affected by the direction of arrival error.In order to exert its characteristics in practical applications,in-depth research on its robustness and complexity is required.In this thesis,a robust adaptive beamforming algorithm for coprime arrays based on vector correlation(RAB-VC)is proposed to improve the robustness of the system.The algorithm in this thesis utilize the correlation between the eigenvectors in the second-order statistics of the coprime array virtual domain and the estimated signal steering vector to reconstruct the interference plus noise covariance matrix,which improves the robustness of the coprime array adaptive beamforming algorithm.Experimental simulations show that the robustness of RABVC make it have a outstanding perfomance in the presence of errors.In the presence of the reaching angle error,RAB-VC has a 10%improvement in the output signal-to-interference-noise ratio compared with the robust beamforming algorithm based on the coprime array,and can also build a deep null at the interference angle.The depth of the beam null is about 10%lower than that of INCM,and the anti-interference ability is stronger.