Research On Three-dimensional Imaging Algorithm Of Interferometric Inverse Synthetic Aperture Radar

Inverse synthetic aperture radar(ISAR)imaging technology has been widely used in military and civil fields because of its ability to generate two-dimensional(2D)high-resolution images of non-cooperative moving targets.However,2D ISAR imaging has some inherent drawbacks.First of all,ISAR images cannot provide the information about the third dimension of the target beyond the range and Doppler dimensions.Secondly,the azimuth dimension of the ISAR image only reflects the Doppler distribution of the scattering points,so the actual transverse dimension of the target cannot be obtained intuitively from the ISAR image.In addition,ISAR images are heavily dependent on image projection planes,that is,the images of the same target formed on different projection planes show strong differences.This brings great difficulties to target recognition.Interferometric ISAR(InISAR)imaging is a three-dimensional(3D)radar imaging technology that combines interferometry and ISAR imaging.It can reconstruct the shape of the target in 3D space without loss of truth,and provide more abundant information for target recognition and classification.Therefore,it has higher research value and wider application prospect.This dissertation concentrates on the practical application of the InISAR 3D imaging,researches and discusses the problems existing in the implementation of moving target InISAR 3D imaging from the aspects of the uniformly accelerated rotation and variable accelerated rotation of the targets,the echo sparsity characteristics and the bistatic system configuration,to enhance the imaging capability of the existing InISAR imaging system and the robustness of the related algorithms.The main research contents of this dissertation include:1.InISAR imaging of uniformly accelerated rotating(UAR)targets.(1)Aiming at the problem that the classical range-Doppler(RD)algorithm cannot perform InISAR imaging on UAR targets,and time-frequency representation algorithms suffer from interference term and large computation,an InISAR imaging algorithm based on peak extraction is proposed on the basis of the RD method.First,the imaging characteristics of the RD algorithm for the UAR targets are analysed,and it is proved that each scatterer still has a peak value in the defocused RD image,and the expression of the peak position corresponding to each scatterer is derived.Then,by extracting the spectral peak of each scattering point and using CLEAN technique,a clear ISAR image is obtained from the defocused RD image.Finally,considering that the ISAR images obtained in this way can retain the essential initial phase information for interference processing,an InISAR imaging method based on peak extraction is proposed.The results of ISAR imaging experiments based on the measured data and the results of InISAR imaging experiments based on the simulated data demonstrate the effectiveness of the proposed method.(2)An InISAR imaging algorithm for the system with non-orthogonal baseline is developed.Firstly,a three-antenna InISAR system model whose baseline is non-orthogonal is established,and a joint motion compensation strategy is proposed to realize motion compensation and range-dimensional image registration simultaneously.Then,by combining the maximum likelihood(ML)algorithm with the fractional Fourier transform(FRFT)algorithm,an ML-FRFT azimuth compression algorithm is proposed.Finally,a coordinate correction technique is proposed to eliminate the deformation caused by the non-orthogonal baseline and achieve the InISAR image of the UAR targets.Simulation results show that the proposed ML-FRFT algorithm can achieve high resolution ISAR imaging of UAR targets,and the proposed coordinate correction technique can effectively correct the deformation of the InISAR imaging results caused by non-orthogonal baseline.2.For the variable accelerated rotating(VAR)targets whose echoes are quadratic frequency modulated(QFM)signal with time-varying amplitude(TVA),an InISAR imaging algorithm based on alternative directional method of multipliers(ADMM)iteration update is proposed.Firstly,the azimuth signal of the VAR target is modeled as TVA-QFM signal.Then,the parameter estimation of TVA-QFM signal is transformed into a convex optimization problem,and an iterative updating algorithm based on ADMM is proposed to solve this convex optimization problem.Finally,the ISAR image of each channel is generated according to the range-instantaneous Doppler algorithm,and then the 3D InISAR images of the target are obtained by interferometric processing.Simulation results demonstrate that the proposed algorithm has superior parameter estimation performance and can achieve the InISAR imaging of the VAR targets under TVA-QFM signal model.3.In order to reduce the adverse effect of envelope alignment error on sparse aperture InISAR(SA-InISAR)imaging,a SA-InISAR imaging algorithm based on iterative optimization for uniform rotating targets is proposed.Firstly,the one-dimensional range profiles of SA data are enveloped-aligned according to the minimum entropy criterion.Then,a gradient based sparse signal reconstruction algorithm is applied to recover the complete signal from the aligned one-dimensional range profiles.By iteratively performing the process of envelope alignment and signal recovery,the signal recovery accuracy and envelope alignment effect can reach the optimal state under certain conditions,and then the optimal recovery signal can be used to obtain relatively high quality ISAR images.Finally,high-resolution SA-InISAR imaging is achieved by interferometric processing of these high-quality ISAR images.Simulation results show that the proposed algorithm can effectively improve the resolution of SA-ISAR and SA-InISAR imaging.4.For the sake of making full use of the advantages of low power consumption,low cost and small size of frequency modulated continuous wave(FMCW)radar and the advantages of anti-stealth,anti-jam and strong battlefield survivability of bistatic imaging,a bistatic FMCW-InISAR imaging method for high speed moving targets is studied by combining FMCW radar with bistatic imaging.Firstly,a bistatic FMCW-InISAR system is designed,and the bistatic equivalent motion model and the corresponding signal model are established.In order to solve the problem of the failure of the "go-stop" hypothesis,a compensation method based on parameter estimation is proposed to achieve range compression.Finally,the conventional algorithm processing flow is combined to complete the InISAR imaging.Simulation results show the effectiveness of the proposed algorithm.