Research On Multiband ISAR Fusion Imaging For Space Target

The progress of the space technology have greatly expanded the breadth and depth of human understanding of the space.As the range of human explore and utilize the space deepening,the space has become a strategic resource and highland.The achievement of modern satellite communications,remote sensing,positioning and other technologies can not leave the support of space satellite,and many threats which influence the national security(ballistic missiles,etc)also come from the outer space.In addition,a variety of space debris are also left in human space exploration,these fragments seriously affect the safety of the spacecraft.These pressing realities urgently require our country to strengthen the abilities of space surveillance and space situational awareness.With its all weather,all time and long range detection capabilities,the inverse synthetic aperture radar(ISAR)play an important role in space surveillance and space situational awareness systems.The high resolution range profile(HRRP)and two-dimensional image provided by ISAR are effective means of target recognition,and the world competing to develop large-bandwidth ISAR to obtain more precise space target images.Multi-band ISAR signal fusion technology expand the bandwidth of radar echo in signal level,and the radar image obtained by this technique are comparable to the real large-bandwidth radar images.Making full use of existing different frequency band radar systems,the hardware manufacturing costs and technical requirements are reduced,so this technology is a valuable approach to obtaining high resolution ISAR images of space targets.In this paper,the key technologies of multi-band ISAR fusion imaging of space target are studied,and the main contents include:1?Multi-band ISAR Echo Modeling and Fusion Condition Analysis for Space TargetThe multi-band echo signals of the same space target are collected from different radar systems,and many error factors(range,time delay,fix phase,etc.)which are not necessary to be taken into account in single band radar may cause the fusion imaging to fail.This paper analysis the signal from the radar signal creation to the echo process systematically,and establish a multi-band ISAR echo model.Based on this,the fusion conditions and multi-band signal phase correction model are proposed,which provide the theoretical basis for the subband signalphase mismatch correction.2?Space Target Multiband ISAR Echo Signal Phase Mismatch Correction ResearchSpace target multi-band ISAR echo signal phase mismatch correction is the basis of fusion imaging,and the mismatch correction results are critical for the signal fusion.First,a minimum entropy and dominant scatterers based phase mismatch correction method is proposed.In this method,the similarity of multiband HRRP and the phase of dominant scatterers are fully utilized,and the signal model do not used here,so the approach has high computational efficiency,which is suitable for large-scale multiband phase mismatch correction.However,the parameter estimation accuracy is affected by Rayleigh resolution.Second,a damped exponent(DE)model and matrix pencil based phase mismatch correction approach is presented.This algorithm,which use the matrix pencil algorithm to estimate the subband signal DE model parameters,solved the mismatch phase by the phases of the corresponding scattering points has higher parameter estimation accuracy.In order to improve the mismatch phase estimate precision,the sparse representation based phase estimation method are put forward in this paper.The subband scattering distribution vectors are obtained by smoothed L0(SL0)algorithm here,after the correspondence relationship between scattering points are established by minimum entropy or cross-correlation,the mismatch phase can be obtained by the scattering points.The local grid subdivision sparse Bayesian learning(LGS_SBL)method are proposed to solve the off-grid problem and to reduce the dictionary dimension.This method contain two steps,the first step determine the scattering location quickly by SL0,then a new dictionary is created after subdivision and selection.the second step solve the Scattering distribution vector by SBL or underdetermined equation solution methods.The parameter estimation accuracy is very high while the computation burden is heavy.3?Space Target Multiband ISAR Fusion Imaging Method and Imaging Resolution AnalysisFor well phase correction multiband signals,a subband frequency shift fusion method is proposed first.With the aid of high parameter estimation accuracy of sparse representation,the frequency band of one subband signal is frequency shift to another frequency which adjacent to the standard subband signal.Only the frequency of one subband signal is changed in this method,and the amplitude characteristics are well preserved,well fusion result can be obtained when the gap band is not very wide.Second,two subband signal are used in the DE model and matrix pencil based iterate fusion imaging method,after the iteration and weighted optimization,the high precision all frequency band predicted data are obtained.Third,the subband signal are equivalent to the sample of all frequency band signal,then a compressed sensing based fusion imaging method is presented.Also,the LGS_SBL algorithm are used to solve the scattering distribution vector,for it requires less manual parameter settings and can work well when there is a strong correlation between the columns of the dictionary.At last,the relationship between the noise,mismatch phase error and the resolution of the fusion images are analyzed.