Research On Space-time Adaptive Processing Algorithm Of Airborne Nonuniform Array Radar

Airborne radar is an indispensable equipment in the fighter and the air early warning,which plays an irreplaceable role in the modern warfare.Due to the non-stationarity of clutter,space-time adaptive processing(STAP)lacks enough effective received data samples,and reduced dimension STAP(RD-STAP)is an effective method to solve this problem.Moreover,limited by the load capacity and space size of the moving platform,the conventional uniform array limits the detection ability of the airborne radar to the target.It is an effective way to solve this problem to make full use of the load size to arrange the array sparsely elements or install elements conformally,combining the polarization radar system and the multiple input multiple output(MIMO)radar system.In the above background,the problem of insufficient samples is more prominent,and conventional reduced dimension processing methods are not suitable for nonuniform arrays.Therefore,for airborne nonuniform arrays,this dissertation focuses on space-time two-dimensional signal processing on how to design reduced dimension algorithms to improve clutter suppression performance.The main work and contributions are as follows:(1)Research on the space-time channel construction method of the airborne nonuniform arrayIn the RD-STAP of nonuniform array,the traditional space-time channel construction method has the problem that the channels are not orthogonal,which leads to the degradation of clutter suppression performance.In order to solve the problem,this dissertation proposes a channel construction method based on the zero point of the pattern.The method utilizes the numerical analysis to find the zero point of the pattern as the frequency point of the spatial channel.The space-time channels can be constructed by combining the time-domain channel.Compared with the traditional channel construction method,the proposed algorithm effectively improves the output signal-to-clutter-and-noise ratio(SCNR)in both joint domain localized(JDL)and best channel method(BCM)RD-STAP scenarios.When the scanning direction changes,there is no need to recalculate the zero points,so the space-time channels can be quickly constructed.(2)Research on reduced dimension polarization-space-time processing algorithm of nonuniform array radar under the condition of small samplesIn the nonuniform array polarization space-time processing,the large signal dimension and the insufficient received data samples lead to the clutter suppression performance degradation.In order to solve the problem,this dissertation proposes a channel selection algorithm based on sparse recovery.A small number of samples are used to estimate the clutter noise covariance matrix by the sparse recovery method.Then the influence of each polarization angle Doppler channel on the output performance is evaluated,and several channels with the greatest impact are selected to improve the performance of the reduced dimension polarization STAP.At the same time,the detection performance of the slow targets is greatly improved.(3)Research on channel selection algorithm in reduced dimension space-time processing for airborne nonuniform array MIMO radarFor the MIMO nonuniform virtual array,the signal dimension is too large and the effective samples are limited.This dissertation proposes an angle-Doppler channel selection algorithm based on the MM framework.Based on the working mode of the MIMO virtual array,the lower bound of the optimal output SCNR is deduced.The optimization problem is designed from the perspective of global optimization.And the problem is transformed into a quasi-convex form using the Schur complement and MM algorithm framework.This algorithm has great advantages when the number of received data samples is limited.(4)Research on the optimization method of auxiliary channels selection in reduceddimension space-time processing of conformal arrayIn the space-time processing of conformal arrays,irregular airspace channels structure and traditional auxiliary channel selection result in the poor performance of STAP.To tackle the problem,this dissertation proposes a channel selection method based on penalized sequential convex programming.With generalized sidelobe cancellation as the processing structure,the blocking matrix,channel selection vector and reduced dimension filter weights are jointly designed.The space auxiliary channels construction,channels selection and STAP weight vector optimization are integrated.Compared with traditional methods,the proposed algorithm effectively improves the output SCNR and is suitable for various non-uniform arrays.