Research On Clutter Suppression And Target Detection Algorithm Of Airborne Radar

Airborne radar faces a complicated clutter environment while detecting targets in ground and sea,space time adaptive processing(STAP)suppresses the clutter effectively in both spatial domain and time domain,therefore the target detecting performance is increased.In consideration of the receiving data of airborne radar is always nonhomogeneous in practice,this dissertation studies the STAP technique and adaptive detection algorithms under nonhomogeneous environment,and the application of STAP technique in multiple-input multiple-output(MIMO)radar and sparse array is further discussed,the main contributions of this dissertation are as follows:(1)The sea clutter modeling of airborne radar is analyzedThe movement of platform leads to the spread of clutter power in spatial domain and temporal domain,the physical mechanism of sea clutter is complex and the behavior of non-Gaussian and non-stationarity are significant,hence modeling of sea clutter is the foundation to the study of signal processing in airborne radar.The clutter echo of airborne moving platform is modeled based on the integrating clutter model,and the scattering mechanism is discussed.Based on research that concentrates on statistical property,this dissertation studies several amplitude distribution models and the corresponding amplitude statistical curves under different parameters.By exploiting the real measured sea clutter,several estimation methods of sea clutter distribution parameters is introduced,then the fitting accuracy of different distribution is compared.(2)The STAP algorithm of nonhomogeneous clutter in airborne radar is studiedWhen the detection environment of airborne radar is nonhomogeneous,the independent and identically distributed(i.i.d)training samplers used for STAP is limited,results in the difficulty for estimating the clutter covariance matrix(CCM)accurately.Three CCM estimation approaches with limited samplers are proposed,which has an accurate estimation performance in the case of few training samplers.Several approaches of selecting auxiliary channels of common reduced dimension(RD)STAP are discussed based on the fundamental principle of STAP,but the fixed auxiliary channels are often non-optimal.In order to select the RD channels adaptively,a new RD STAP method is proposed though the mathematic operation of the output SCNR in transform domain,which according to the angle-Doppler correlation coefficient,and then extend the situation to preserving the whole spatial degrees of freedom.Finally,the efficiency of algorithm is verified by numerical simulation and measured data processing.(3)Study the knowledge-aided adaptive detection algorithm in sea clutterThe ultimate purpose of STAP is detecting target,space time adaptive detection for airborne radar realizes the combination of clutter suppression and target detection.Aiming at the sea clutter background simulated by compound-Gaussian distribution,the generalized likelihood ratio test(GLRT)for detecting point target and range spread target is studied with the texture component satisfies two parameter inverse Gaussian distribution.According to the priori distribution of texture component and speckle component,the detection performance of different adaptive detector is compared based on knowledge-aided CCM estimation methods.We address the moving target detection problem under GLRT framework without using any training sample signals with radar placed on moving platform,through utilizing the low rank property of CCM to estimate the amplitude of clutter basis and noise component,results in better detection performance comparing with the GLRT detector depending on training samples in case of limited training signals.(4)Airborne MIMO radar STAP problem studyAiming at the application of MIMO radar in airborne platform,the range sidelobe suppression technique of frequency divided linear frequency modulation(FDLFM)signal is studied,via modifying the spectrum and adding window to frequency domain of composite MIMO radar signal.The wideband effect should be taken into consideration for signal processing when the frequency interval between each orthogonal signal becoming larger,this dissertation presents the wideband frequency divided signal model and the estimation formula of clutter rank in case of wideband MIMO radar,a pre-filtering method is proposed to eliminate the Doppler frequency shift caused by the increase of frequency interval,simulation results show that the detection performance of high speed target will be improved effectively.According to the three-dimensional cascaded structure of MIMO radar,a signal processing framework is proposed by combining signal data from each stage and disposing different filters in the corresponding stage flexibly,the comparison of computational cost and clutter suppression performance of several typical methods are compared.(5)The STAP algorithm of sparse distributed antennas and pulses is studiedBesides MIMO radar,the DOF can be extended through non-uniform placed antennas with specific structure as well.This dissertation introduces the signal model of nested array and outlines the configuration of several typical sparse sensor arrays.Bringing in the concept of nested array into the antenna layout problem,we arrange the antenna geometry and temporal sample interval to make the subset of joint space time samples satisfy the nested samples,and reconstruct the CCM via the structural information imposed on CCM,leads to a significantly enhance of space time DOF.In addition,consider an airborne radar with uniform linear array(ULA)and transmits coherent pulses,we select the antenna pulse pairs which satisfy the location of nested samples and estimate the compressed CCM,then an antenna pulse selection strategy is proposed based on maximizing the output SCNR.Proposed algorithm has lower computational cost and better selection strategy compared to previous methods.