Research On Imaging Of Low-oversampled Staggered Synthetic Aperture Radar

Synthetic Aperture Radar(SAR)imaging is unable to achieve high resolution in azimuth and wide swath in range,simultaneously.To solve this problem,researchers proposed an innovative concept——Staggered SAR using the variable Pulse Repetition Interval(PRI),which can effectively achieve High Resolution and Wide Swath(HRWS)imaging.Low-Oversampled Staggered SAR(LS-SAR)refers to the Staggered SAR adopting the azimuth oversampling ratio between 1.1 and 1.5,which reduces the data storage requirements and the cost of hardware implementation,leading to the important research significance.However,low azimuth oversampling ratio,echo data loss and nonuniform sampling will result in severe azimuth ambiguities in the final focused image,that is a large number of dense and high-energy false targets in the azimuth direction of the true targets.In order to suppress the ambiguities for LS-SAR,this thesis studies the two-step ambiguity-removal method based on Missing Data Iterative Adaptive Algorithm(MIAA).Moreover,an ambiguity-removal method based on Compressive Sensing(CS)is proposed.The main work of this thesis is listed below:1.The working mechanism of the Staggered SAR system to achieve HRWS imaging is introduced,and its echo model and PRI variations modes are studied.Then,the effects of low oversampling ratio,echo data loss,nonuniform sampling and the PRI variation mode on azimuth ambiguity in LS-SAR are studied.2.To suppress the severe ambiguities in LS-SAR image,the two-step ambiguityremoval method based on MIAA is introduced.First,to solve the problem of sub Nyquist sampling caused by echo data loss in LS-SAR,the MIAA recovery method is introduced.Second,to solve the nonuniform sampling in azimuth,the Nonuniform Fast Fourier Transform(NUFFT)or the Conformal Fourier Transform(CFT)can be utilized.Finally,simulations verify that NUFFT has a higher tolerance to low oversampling ratio than CFT does,and MIAA is effective for point targets but cannot completely suppress the azimuth ambiguity in distributed targets due to its non-sparse spectrum in azimuth.3.Considering the deficiency of the MIAA-based method that the ambiguities in distributed target cannot be effectively suppressed,an ambiguity-removal method based on CS for LS-SAR imaging is proposed.First,a Hybrid-Domain Model(HDM)based on the relationship between the echo data and the scene image is constructed,which accurately describes the coupled range cell migration,the nonuniform sampling and the echo data loss.Then,based on the HDM,an optimization problem is constructed,which is then solved by the 2D Fast Iterative Shrinkage Thresholding Algorithm(2D-FISTA)with Fast Fourier Transform(FFT)and NUFFT to get high-quality imaging result,efficiently.Simulations and the simulated raw-data experiments demonstrate the effectiveness and efficiency of the proposed method over other existing methods.