Spaceborne Video SAR Mode Design and Data Processing

‍ ‍The video Synthetic Aperture Radar （SAR） can achieve continuous observation of key areas and present them in a video format. Its high frame rate imaging capability is advantageous for real-time detection of ground and sea-based maneuvering targets. Video SAR and its related technologies have become a research hotspot recently as a novel SAR observation mode. Starting from the principles of synthetic aperture， the higher the SAR carrier frequency， the larger the absolute value of azimuth frequency modulation rate， leading to shorter synthetic aperture time for achieving the same resolution， which is conducive to video generation. However， the higher SAR carrier frequency lead to the larger attenuation of signal transmission. After considering the compromise， the Ka band was selected to realize the spaceborne video SAR. Through a well-designed system， Ka-band space-borne SAR can work on video mode. However， space-borne SAR operates at a long range， and there are significant differences in the geometric model， waveform design result compared to airborne SAR for video observation. Furthermore， space-borne video SAR has a wide swath， requiring large data volumes and significant computational processing. Fast data processing is a crucial consideration for timely image acquisition and subsequent applications. This paper conducts relevant research based on the Luojia-2 SAR payload and measured data. Firstly， a practical video mode waveform is designed based on the Luojia-2 SAR satellite orbit and platform capabilities. Since the orbit of the satellite is ellipse， it is hard to realize long time observation and circular SAR. Spotlight mode is adopted. Meanwhile， since the hardware of SAR payload can work on the variation of PRF （Pulse Repetition frequency）， sampling start time and sampling data length mode and it benefits to less data rata， which is adopted in waveform design. Secondly， considering the highly efficient feature of frequency domain imaging algorithm， it is an alternative algorithm in data processing. However， the imaging result of frequency domain imaging algorithm depends on imaging geometry， which is different for different look angle， an extra coregistration procedure is required and lead to low efficiency and large error， a pipeline processing procedure suitable for rapid imaging of space-borne video SAR based on time-domain imaging algorithms is proposed. Benefits from the high efficiency of CFBP （Cartesian Factorized BP） algorithm and pipeline processing with GPU （Graphics Processing Unit） acceleration， it can achieve fast data processing for space-borne video SAR. Finally， video SAR imaging processing is completed based on the measured data of Luojia-2. From the imaging results， a cargo ship is clearly observed near the river， which is moving close to the shore. This experiment realizes the acquisition and imaging processing of space-borne Ka-band video SAR data for the first time， demonstrates the real-time detection capability of space-based video SAR on ground/sea surface maneuvering targets， and the imaging result is clear and complete， which verifies the correctness of the proposed spaceborne video SAR fast imaging algorithm. The shortcomings and improvement directions of this study includes the hardware （DSP and FPGA） implementation and real-time processing development of the algorithm.