Research On Imaging Techniques For High-Resolution UWB SAR Systems

High-resolution is an important development direction of Synthetic Aperture Radar(SAR).Ultrawideband(UWB)SAR can achieve range resolution better than 0.3m or tiwce the wavelength,which has the advantages that conventional narrowband SAR cannot match in high-resolution imaging,target recognition and penetration.In addition,the combination of UWB and beam steering technology enables SAR to obtain two-dimensional(2D)high-resolution images at or better the decimeter level.However,the large range bandwidth and the wide azimuth beam angle make the range-azimuth coupling serious,which brings challenges to the classical SAR signal processing theory.This paper focuses on the technical issues related to the imaging processing of high-resolution SAR systems in decimeter level,and solves the problems of accurate compensation of high-order coupling terms in 2D spectrum and azimuth spectrum aliasing.We hope that the theories and conclusions in this article can provide a support to the real high-resolution SAR systems.The highlight points and innovations of this paper are as follows:In Section 2,an improved generalized chirp scaling(GCS)algorithm based on Lagrange inversion theorem is proposed to solve the serious 2D coupling problem of low frequency UWB SAR data.The large range bandwidth and wide beam make the range-azimuth coupling of the echo signal serious.Especially in the low frequency SAR,the coupling property is more obvious.Thus,the existing frequency domain approximation algorithms are no longer applicable.The improved GCS algorithm can effectively compensate the high-order terms by high-order range-independent coupling phase compensation and Lagrange inversion.The proposed algorithm can be widely applied to imaging of SAR systems such as high resolution and multi-mode.In Section 3,there are three kinds of time-frequency transform methods:two-step technique,fractional Fourier transform and moving band chirp-z transform(MBCZT).This paper proves the equivalence of the three methods.The MBCZT-improved GCS(MBCZT-GCS)algorithm is proposed to process UWB beam steering SAR data.On the one hand,the MBCZT is used to solve the azimuth spectrum aliasing problem.On the other hand,the improved GCS algorithm is used to solve the 2D coupling and achieve efficient focusing.In Section 4,the signal properties of beam steering SAR in UWB and squint are analyzed,and an imaging processing method combining MBCZT and squint ? K is proposed.Compared to the broadside beam steering SAR,the coupling is more serious in the case of squint.In addition,due to the range frequency dependence of the Doppler center,the azimuth bandwidth composition of the squint is different from that of broadside.The azimuth total bandwidth is composed of the beam bandwidth,the beam steering bandwidth and the squint bandwidth.The azimuth instantaneous bandwidth is composed of the beam bandwidth and the squint bandwidth.Under the premise that the PRF is greater than the instantaneous bandwidth,the MBCZT can solve the azimuth aliasing problem.The squint ? K uses the modified Stolt interpolation kernel,which is more effective for squint data focusing than the Stolt interpolation.