Research On Key Technologies Of Ultra-high Resolution Airborne SAR Imaging

Synthetic Aperture Radar(SAR),as one of the most important technologies in remote sensing,has shown great value in many fields for its ability to achieve fine-resolution images of targets in all-day and all-weather conditions.On one hand,Ultra-high range resolution SAR imaging requires ultrawideband LFM signal,which can be acquired by adopting stepped chirps technology.However,traditional synthetic bandwidth technique could not completely avoid phase error during pulse synthesis.On the other hand,it's hard for traditional stripmap SAR to achieve wide azimuth beam due to the limitation of antenna size,resulting in the difficulties to obtain high azimuth resolution and multi-aspect information of target.In addition,in order to ensure the quality of ultra-high resolution MiniSAR imaging,it's necessary to develop the traditional autofocus algorithm to restrain the phase error which is variant in two dimensions.Also,the parallel processing solutions can be designed to improve data processing efficiency with GPU and CUDA.Based on the problems above,the main work of this paper are as follows:1.Based on the waveform model of stepped chirps,the classic time-domain and frequencydomain bandwidth synthesis methods with sub-pulses compensation within a burst are investigated.Nevertheless,some minor phase is inevitable during pulse synthesis.To eliminate this disadvantage,two direct bandwidth synthesis methods based on PFA and Omega-K algorithm are studied with detailed derivation and explanation,which embed the error compensation between channels into the imaging algorithm.Finally,a detailed real-data processing flow is presented with PGA technology to compensate for the error in sub-bands and between sub-bands.2.In the traditional stripmap SAR,it's hard to achieve wide azimuth beam due to the limitation of antenna size.To solve the problem,this paper proposes a scheme for 90° wide-angle stripmap SAR imaging based on beam scan.First,the method based on beam scan of the phased array,the selection and combination of beam position pulses are described for wide-angle SAR imaging in stripmap mode.Then,the limitation for the number of beam position and radar parameters is investigated.In addition,the approach based on the equal segmentation of the synthetic aperture is discussed to obtain the SAR imagery of the targets under different squint angles.The simulation results based on BPA verify that the proposed scheme for wide-angle stripmap SAR imaging can effectively obtain high azimuth resolution and the multi-aspect scattering information of target.3.The Omega-K Algorithm based on dechirp signal is derived for ultra-high resolution MiniSAR imaging.The motion compensation technology is explained to avoid image deterioration resulted from severe jitter of aircraft.Also,in order to effectively correct the phase error of MiniSAR,which is variant in two dimensions,an improved range segment autofocus technology is proposed.Finally,we design a GPU parallel acceleration scheme based on the imaging algorithm framework,which can efficiently obtain ultra-high resolution MiniSAR images with high quality.