Adaptive Interference Suppression And Knowledge Aided Space Time Signal Processing For Airborne Radar

The aircraft equipped with phased array radar can be deployed flexibly and rapidly to detect and track various targets in the battlefield,perceive the enemy's situation,and provide intelligence support for the command and control of its own forces.However,there are still several problems remain to be clarified,such as the interference suppression and clutter suppression.In the part of signal processing,adaptive beamforming is usually used to suppress the sidelobe interference.When the array model is inconsistent with the actual situations,the SINR output performance will be reduced.Random direction errors,array position mismatches,amplitude/phase errors and local coherent scattering will affect the performance of the adaptive beamformer.Therefore,it is important to design a robust beamformer to reduce the impact of the errors.On the other hand,as the platform of airborne radar is moving,the ground clutter shows space-time coupling characteristics,it is necessary to use space-time adaptive processing technology to suppress the ground clutter received by airborne radar.However,due to the influence of radar antenna configuration and external working environment,the received echo signal is usually inhomogeneous,which can not provide enough independent and identically distributed sampled data.How to improve the performance of STAP in inhomogeneous environment is also the main problem to be studied.This paper focuses on the above issues,including the following aspects.1.A robust adaptive beamforming algorithm based on sparsity is proposed for adaptive beamforming with few samples.When the number of samples is small,the estimation accuracy of sampling covariance matrix deviates from the ideal interference plus noise covariance matrix.The traditional diagonal loading method and the worst-case method are robust for the case of small samples.However,when the sample data contains the target signal,the performance of the two methods is seriously reduced.To solve this problem,a robust adaptive beamformer based on gridless spice is proposed.Compared with other methods,when the number of samples is small(or single snapshot),the performance of this method is closest to the optimal performance,and it is robust to the expected signal angle mismatch and coherent local scattering error.In order to solve the model mismatch problem,a robust adaptive beamformer based on ADMM is proposed.The influence of model error(array amplitude and phase error or array spacing error)can not be ignored.For the robust adaptive beamformer,which assumes that the array configuration is known,model mismatch degrades their performance.This paper presents a robust adaptive beamforming technology based on ADMM.The simulation results show that the method is robust to the angle error,amplitude and phase error,space error and coherent local scattering.2.Combined with the research of prior terrain data and geographic elevation data,a STAP method based on prior information is proposed.Based on the convex combination of sampling covariance and prior covariance,this paper proposes a prior information STAP method.Firstly,the prior covariance matrix of the simulation is obtained by using the surface classification data and geographic elevation information,through coordinate transformation,occlusion judgment and data simulation preprocessing.In order to estimate the sampling covariance matrix more accurately,the regularized Tylor estimator is used.The experimental data processing results prove the validity of the method.3.To solve the problem of sample selection in inhomogeneous environment,a method of sample selection based on prior information is proposed.The training samples used to estimate the STAP clutter covariance matrix in inhomogeneous environment may have different distribution characteristics from the clutter of the unit to be processed,which will lead to the inaccuracy of the estimated clutter covariance matrix and the degradation of STAP performance.In this paper,a training sample selection method based on prior clutter covariance matrix is proposed.Firstly,a prior covariance matrix is used to detect the training sample containing the target signal.Then the contaminated sample is projected into the subspace constructed by the prior covariance matrix to retain the training sample.Finally,the final clutter plus noise covariance matrix is obtained by reweighting the training samples with iterative method.Considering the characteristics of the cells to be processed,this method can accurately estimate the clutter covariance matrix of the cells to be processed,which can effectively improve the performance of STAP in non-uniform environment.4.A gridless space-time adaptive processing method based on the minimization of total variation is proposed for clutter suppression in non-stationary environment.Due to the influence of radar antenna configuration or external environment factors,the clutter spectrum is non-stationary in distance,which results in the number of training samples not meet the conditions of independent and identical distribution and affects the final clutter suppression performance.To solve this problem,a space-time adaptive processing method based on the minimization of total variation without grid is proposed.Using the sparsity of clutter in the space-time plane,an optimization problem based on the atomic norm is established,and the two-dimensional optimization problem is transformed into one-dimensional one by using the polar coordinate representation of the guidance vector and Bessel function approximation.Finally,using the properties of Toeplitz matrix,CVX toolbox is used to solve the optimization problem.Compared with the traditional sparse processing method,this method has better performance in sidelooking and non sidelooking cases.