Study Of Robust Adaptive Beamforming Algorithms For Antenna Array

Adaptive beamforming technology is one of the important branches in array signal processing,and an effective method for adaptive beamforming used for enhancing a desired signal while suppressing noise and interference in the spatial signal processing.It is widely used in both military and economic fields,such as communication,radar,navigation,sonar,radio astronomy,acoustics and speech signal processing,medical imaging,and other areas.Aadaptive beamforming technique becomes more and more mature in theories,but still many practical issues in engineering applications.There are various errors in the actual system,which will directly affect the effectiveness of adaptive beamforming algorithm in practice.Therefore,it has been a popular topic that how to enhance the robustness of adaptive beamforming algorithms.This dissertation mainly studies the robustness of adaptive beamforming technology.The main research contents are summarized as follows.For the problem of the standard Capon beamformer,by making the projection of signal steering vector onto noise subspace optimized,a new robust Capon beamforming algorithm based on steering vector uncertainty set is proposed.The noise subspace is obtained by the feature decomposition of the array covariance matrix,and utilizing the feature of characteristic vector,an solution of the optimal diagonal loading is derived.For the robust adaptive beamforming problem of the general signal models,a robust adaptive beamforming algorithm based on the worst-case performance optimization with positive semidefinite constraints is proposed.A more simple robust adaptive beamforming algorithm is derived by transforming the original beamforming algorithm.Solving the beamforming algorithm,an approximate closed-form solution to the adaptive weight vector and some improvement of the performance are derived with low complexity.To solve the decrease of the performance in the adaptive beamformer in the situation of limited sample size,this paper presents a robust adaptive beamforming algorithm based on the covariance matrix estimate.This algorithm obtains an enhanced covariance matrix estimate via a shrinkage method,instead of the traditional sample covariance matrix without any parameters.In order to overcome the influence of the performance of the adaptive beamformer in the presence of array steering vector errros,a robust adaptive beamforming algorithm based on mismatch errors orthogonal decomposition is proposed by using the orthogonal component of the errors to correct desired steering vector without setting any parameters.By combining the two algorithms obtain a new hybrid robust algorithm which is robust against covariance matrix uncertainty and the signal steering vector mismatch.In the multipath environment,to overcome the influence of array mismatch errors on the performance of spatial smoothing algorithm,a robust adaptive beamforming algorithm based on the spatial smoothing is proposed.The algorithm is based on worst-case performance optimization criterion,which ensures that all possible in the desired signal steering vector and sample correlation matrix maintain a distortionless response while minimising the array output power.A non-spatial smoothing algorithm for robust adaptive beamforming is proposed with the assumption that the directions of the correlated interferers have been estimated in advance.The algorithm is also based on worst-case performance optimization criterion,the desired signal with steering vector mismatch is maintained distortionless at the output by the desired signal constraint.On the other hand,the coherent interferences are suppressed with coherent interferences constraints and the uncorrelated interferences are suppressed by minimising the output power.Further improvement of the output signal-to-interference-puls-noise ratio(SINR)for the beamformer can be achieved by removing the desired signal in the received data with a subtraction operation,the optimum weight coefficients are obtained form the new data vector.Based on the time-domain broadband array structure,a robust adaptive broadband beamforming algorithm with magnitude response constraints is proposed.The algorithm can precisely control the mainlobe shape and effectively overcome the influence of signal look direction mismatch by using magnitude response constraints.Simultaneity,the algorithm utilizes the spatial response variation constaint to design the frequency invariant beampattern and improve the frequency invariant property of beampattern.For the broadband array structure,a robust adaptive broadband beamforming algorithm based on worst-case performance optimization is proposed.The algorithm utilizes spatial response variation constaint to imprve the frequency invariant property of beampattern?In order to improve the robustness against the steering vector mismatch,the uncertainty set constraints of the steering vector and spatial response variation are used,finally a closed-form solution is obtained.