Research On Ground Moving Target Detection And Parameter Estimation In The Image Domain For Multichannel SAR

Multichannel synthetic aperture radar(SAR)ground moving target indication is a important function for the modern SAR system,which has been widely used in ocean current measurements,traffic monitoring,battlefield reconnaissance and so on.In the SAR image domain,the ground moving target has a high signal-to-clutter-noise ratio,and its detection result is more visible and convenient for subsequent identification.However,the focused strong static target in the image domain is hard to suppress completely and may cause false alarm.Furthermore,in the multichannel SAR,the target's echoes are sampled first by the pulses in the time domain,and then sampled by the channels in the space domain,which may lead to a cascaded time-space Doppler ambiguity(CTSDA)problem when the target's velocity is large.The space domain Doppler ambiguity(SDDA)can cause the azimuth location ambiguity,and the CTSDA will seriously degrade the velocity estimation performance.Therefore,the thesis focuses on the difficulties in the multichannel SAR ground moving target detection,location and velocity estimation in the image domain.The main work of the thesis can be summarized as follows:1.The multichannel SAR ground moving target's signal model is derived first in the image domain,then some target's properties in the image domain are quantitatively analyzed,such as the CTSDA,energy variation,azimuth shift and second-order phase modulation,which lay a foundation for the multichannel SAR ground moving target indication and parameter estimation in the image domain.2.To overcome the false alarm caused by the residual strong static target after clutter suppression in the image domain,a ground moving target detection method based on relative residue of DPCA(RR-DPCA)is proposed.The multilook images of dual channels are calculated to normalize the original DPCA outputs,which makes the RR-DPCA be more sensitive to the moving target velocity than DPCA.Furthermore,the homogenous clutter statistical distribution after RR-DPCA is analyzed,then relative residues but not the absolute residues are used for the subsequent CFAR detection.Finally,both the simulated and real data are provided to demonstrate the RR-DPCA method can well suppress the strong static target.3.To solve the ground moving target's azimuth location ambiguity problem caused by the SDDA,a co-prime arrays SAR(CA-SAR)is proposed first,where two sparse uniform linear subarrays with co-prime element spacings are utilized to construct a virtual array with a small element spacing,which can increase the space domain sampling frequency and solve the SDDA as well as the azimuth location ambiguity.However,the CA-SAR requires at least one physical element spacing to be small enough.Then a sparser co-prime adjacent arrays SAR(CAA-SAR)is proposed by changing the arrangement of the two subarrays.Based on numerical experiments and performance analysis,it is validated that the proposed CA-SAR and CAA-SAR have a better parameter estimation performance than the classic nonuniform linear antenna array SAR.4.To overcome the CTSDA problem of the radial velocity estimation based on the interferometric phase,a multifrequency SAR is proposed.The CTSDA with multifrequencies can be modeled as a double remaindering problem,and a sufficient condition for the uniqueness of a solution is derived.When the sufficient condition is satisfied,the double remaindering problem can be degenerated to the conventional congruence problem and solved by the existing closed-form robust Chinese remainder theorem.Otherwise,a searching based reconstruction method is proposed.Finally,some results of numerical experiments are provided to demonstrate the effectiveness of the proposed methods.