Robust Adaptive Beamforming Algorithm Of Anti-interference Array Antenna In Complex Environments

As the important research direction in the field of array signal processing,adaptive beamforming technology is widely used in military defense and people's livelihood development because of its efficient ability to suppress interference and preserve the desired signal.Based on such technology,the array antenna can adjust the weight coefficient of each array element to steer the main lobe of the radiation beampattern towards the desired signal while place respective nulls towards several interference for achieving the high output signal-to-interference-plus-noise ratio(SINR)performance of array antenna.Adaptive beamforming possesses the superior spatial filtering ability of nulling electronic interference under the ideal non-complex scenarios.However,in practical engineering applications,it is inevitable to face the adverse effects of various model mismatch errors and unusual noise interference in complex environments,which make some of the beamforming algorithms suffer from performance degradation or even fail.Therefore,how to improve the robustness of the beamforming algorithms toward the non-ideal factors in complex environment,that is,to reduce the sensitivity to intentional or unintentional mismatch errors and interference noise,is an urgent problem to be solved in current engineering applications.This dissertation summarizes the classic beamforming criterias and conventional beamforming algorithms,and then focuses on the in-depth research on the four aspects of mismatch error factors in non-ideal application scenarios including impulsive noise interference,high amplitude of sidelobe and low level of nulling depth,steering vector mismatch of the desired signal and the spatial overlap interference,and the corresponding robust adaptive beamforming(RAB)algorithms are proposed.The research contents and contributions of this article are summarized as follows:1.We work on the problem of the severe performance degradation of adaptive beamforming algorithms under the impulse noise interference environments.Two RAB algorithms based on the correntropy criterion and the norm modulus regularization are proposed.In order to improve the robusness against impulsive noises,the correntropy criterion is introduced in the beamforming algorithms for its insensitivity to external outliers.The algorithms also take into account the possibility for provoking sparsity of L1-norm to the weight coefficients that force the feedback current of the array element to zero.Then,we utilize the convexity of L2-norm to make a convexification transformation to the objective function of weight solution so as to force the flat optimized regions of objective function to further curve upward,and improve the accuracy of the weight solution.The algorithms embedded in adaptive beamformer system could make the device possess the superior ability of robustness against impulsive noise and energy-saving for the antenna array with a number of active feedback elements.2.The beamformer systems with the problem of high sidelobe level and insufficient null depth of the beampattern for iterative optimized and derived-based algorithms are studied.The RAB algorithms based on the Fibonacci branch search(FBS)optimization strategy are proposed.The proposed FBS algorithm is an extensible multi-dimensional objective optimization method based on the one-dimensional optimization strategy with the Fibonacci sequence.The interactive iterative rule and the global random search mode are adopted in the search elements to make the FBS algorithm have the global optima search ability for multi-dimensional target optimization problems.In the framework of adaptive beamforming(ABF)model,the global optimum weight is searched by FBS in terms of the fitness function designed on the basis of the interference nulling depth and the sidelobe level,achieving low side lobe level and high interference depth null of the beamformer.3.The dissertation engages in the problem of the performance degradation of the beamforming algorithms caused by the steering vector mismatch for the desired signal.Two RAB methods are proposed to resist the steering vector mismatch of the desired signal.The first method utilized the generalization matrix to perform high-order reconstruction of the interference and noise covariance for reducing the self-cancellation of the desired signal component.Then,the strategy of spatial domain spectrum segmentation integral is used to construct the model uncertainty set for the steering vector of the reconstructed desired signal to accurately estimate the arrival angle of the desired signal.The second method takes the constrained programming condition of the constrained mean square error algorithm as the framework of the optimization model,and the steering vector mismatch of the desired signal is introduced into the constrained equation to construct a non-convex constrained optimization problem.Combine the closed-loop convex optimization conversion and constrained stochastic gradient to solve the constructed optimization problem,and improve the robustness of the algorithm against the steering vector mismatch of the desired signal.4.The performance degradation of the optimal adaptive beamforming algorithm caused by spatial overlap interference is got into research.Two RAB algorithms are proposed to deal with the spatial overlap interference caused by the wavefront distortion.The first algorithm takes advantage of the karhunen–loeve transformation to extract the spatial processing matrix based on the spatial overlap interference covariance.The robustness of the beamformer against such interference is improved by weakening the spatial overlap interference based on the spatial processing matrix.The second algorithm utilized the conversion property between the high-dimensional toeplitz matrix and the circulant matrix to acquire the equivalent spatial processing matrix in the form of discrete fourier transform matrix for the spatial overlap interference.The adverse effects of the spatial overlap interference to the beamformer are weakened by means of the filter processing implementation based on the extracted spatial processing matrix.Finally,the ability of the proposed algorithm to suppress spatial overlap interference is analyzed from the aspect of quantity and quality.