Research On Methods Of Sparse Apertuer ISAR/InISAR Imaging

Inverse synthetic aperture radar(ISAR)has the advantage of all day,all weather,and observe the target from a long distance.Usually,the imaging result of the observed target from the ISAR system has high resolution.The latter automatic target recognition(ATR)processing can be completed based on the feature information extracted from the ISAR imaging result.Thus,ISAR iamging has indisputable benefits in both military and civilian applications.As the radar operating system becomes multi-functional and the observed scene becomes complex,modern phase array radar system usually nee d to carry out multiple tasks with the help of time control mechanisms.To finish tracking,recognizing,imaging tasks of the radar system,the direction of radar beams sometimes have to switch continuously among different observed targets in a short period of time.In addition,deliberate interference and the complex motion of the observed target can also lead to the missing of echoes.Thus,a single target cannot achieve the consecutive illumination of the radar beam,which results in the sparse aperture(SA)situation of the echo data.As the coherence of echoes descends,traditional motion compensation method becomes ineffective or lose efficacy.Make zero padding to the incomplete data and adopt range Doppler(RD)method will produce evident interference stripes in the imaging results,which is adverse to the latter ATR processing and classification.As the target’s pose and mode of motion in different imaging time and image projection plane(IPP)produce different ISAR images,which also brings trouble to ATR processing.Therefore,considering the problem of motion compensation and ISAR imaging under sparse apertures observation situation,ISAR imaging method and In ISAR imaging method under SA condition are studied by utilizing minimum theory and compres sive sensing theory,as well as spatial spectrum estimation technique and interferometric ISAR(In ISAR)imaging technique in this paper.The main contribution of this dissertation includes the following four parts:1.To process the echo signal in the case of SA situation,firstly,we establish the signal model under SA situation,then entropy minimum based range alignment method and eigenvector based phase adjusment method are employed to realize motion compensation of echo signals.Furthermore,gradient desc end based method is proposed to recover the missing part of the echo data and form the ISAR image under SA situation.Finally,simulated data and real measured data experiments show that the proposed methods can accomplish the motion compensation and signal reconstruction effectively,and obtain the well-focused ISAR imaging results.2.To sovle the problem of ISAR imaging under SA observation situation,a two-dimensional unitary matrix pencil(UMP)method is proposed to form the ISAR image of the observed target,reffering to the spatial spectrum estimation technique.Firstly,Gerschgorin disc estimator(GDE)method is utilized to estimate the number of scatterers in the target.Secondly,two-dimensional UMP method is employed to estimate the two-dimensional positions of scatterers in the target.Then,the least squares(LS)method is used to estimate the amplitude information of the scatterers in the target.Based on the position and amplitude information of scatterers obtained above,the scatterers can be projected onto the image plane to obtain the ISAR imaging result.Considering the computational load of the two-dimensional UMP method is huge,the total least square-estimating signal parameter via rotational invariance technique(TLS-ESPRIT)method is proposed to estimate the locations and amplitude information of scatterers.Finally,the simulated date and real measured data experiments validate the proposed methods can obtain satisfactory ISAR imaging performance under SA observation situation.3.To solve the problem of data missing under SA observation situation,a message passing based method is proposed to realize the ISAR imaging,refferring to the compressive sensing theory.First,the prior probability desnity function(PDF)of the signal and the statistical distribution of the noise is given based on the sparse signal model.Then,the Turbo frame is adopted to iterate the variable information in a loop and accomplish the linear minimum mean sqaure error estimation and the posterior estimation.The expectation maximizaiton update strategy is utilized to update the relative statistic parameters in the iteration processing.The simulated data and real meausred data experiments verify the proposed method can achive excellent ISAR imaging results under SA observation condition,and the proposed method is robust to different sparsity rates.4.As the coherent processing interval(CPI)is usually short and the IPP is varying,bi-static configuration and interferometric processing technique are combined to study the three dimensional imaging of ship targets under the SA observation situation.First,the three-dimensional echoes model of the ship target with bi-static configuration is constructed.Then,the coordinate transformation of the scatterers in the target is clarified.To solve the problem of the missing data,alternating direction method of multipliers(ADMM)method is proposed to recover the signal and accomplish the three dimensional imaging.Considering Fourier transformation can bring the effect of widen main lobe and high side lobes,a spatial variant apodization-multiple orthogonal least squares(S-MOLS)is proposed to reconstruct the whole signal,reduce the effect brought by the high side lobes and form the three dimensional image.Simulated data and real measured data experiments confirm that the proposed methods can obtain high quality three dimensional images of the ship target with bi-static configuration under SA observation condition.