Research On Adaptive Beamforming Algorithms Under Nonideal Conditions

Adaptive beamforming algorithm is widely used in wireless communications,speech processing,radar,sonar,medical imaging and other fields.But in practical application,the various existence error sources,like non-ideal position of the array antenna,mutual coupling between array elements,inaccurate angle estimation,low sampling snapshots,colored noise,high dynamic environment and so on,effect the output performance of the beamformer and even cause serious performance failure.Therefore,how to improve the performance of beamforming algorithm in non-ideal environment has become the focus of many researchers.In this paper,the basic theoretical knowledge of beamforming algorithm is briefly summarized,and the influence of common nonideal environment on beamforming algorithm is analyzed,and then the corresponding improvement methods are put forward.Specific work is as follows:Limited snapshots and colored noise lead to the waveform distortion of adaptive beamformer.Aimed at solving this problem,two kinds of correction algorithms are proposed: beamforming algorithm based on the constraint Kalman filter in colored noise environment and doubly robust adaptive beamforming algorithm.In the first proposed algorithm,Kalman filter is applied to the field of beamforming filtering,which greatly improves the convergence rate of the algorithm,the state equation and measurement equation of the filtering algorithm are extended by modeling colored noise to eliminate the influence of colored noise on the adaptive beamforming algorithm;in the second proposed algorithm,the doubly beamforming algorithm modifies the covariance matrix by the adaptive shrinkage factor method and then estimates the colored noise by using the modified covariance matrix to eliminate the influence of colored noise on the performance of the beamforming algorithm.Beamforming algorithm based on the constraint Kalman filter in colored noise environment has faster convergence rate and the output SINR of the algorithm is significantly improved under colored noise environment,but the algorithm is more complex,while the doubly robust adaptive beamforming algorithm still has better output performance with lower complexity.The desired steering vector angle mismatch leads to the adaptive beamforming algorithm cannot form the main beam in the desired signal direction,severely the desired signal cancellation phenomenon will occur.Magnitude response constraint is applied to beamforming field to improve the output performance,but the limited snapshots lead to the sidelobe gain level higher.In order to solve the problem above,a robust beamforming algorithm based on magnitude constraint and sparse constraint is proposed.This method takes advantage of the sparsity of the array antenna and adds the sparse constraint in the cost function to ensure that the desired signal is undistorted in desired direction while the gain in the others angles is zero.The magnitude response constraint and sparse constraint are simultaneously added to the cost function of the beamformer,which resulting in receiving the array signal without distortion and reducing the sidelobe level of the beam,make sure that the algorithm has high output performance against a large angle mismatch.Theoretical analysis and simulation results verify the effectiveness of the proposed algorithm.Under normal circumstances,the received data of the array contains the desired signal,using the received data covariance matrix directly to obtain the weight vector causes the beamforming algorithm to suppress the high-power desired signal as an interference signal and futher disturbs the desired signal.Aiming at this problem,the paper proposes an efficient beamforming algorithm for the covariance matrix reconstruction.The algorithm introduces the determination factor as the criterion of the covariance matrix reconstruction.The determination factor is defined by whether the desired signal angle can be estimated by the power spectrum estimation algorithm.Firstly,the power spectrum estimation algorithm is use to estimate the desired signal angle direction,then whether the covariance matrix reconstruction or not is determined by the determination factor,that is,there is no need to reconstruct the covariance matrix in the low SNR,while covariance matrix reconstruction method is adopted in the high SNR,finally the weight vector can be calculated.In this way,the complexity of the algorithm can be reduced while the performance of the algorithm is guaranteed,simulation results show the effectiveness of the proposed algorithm.