Study On High Resolution/Wide Swath SAR Imagery And Motion Compensation Algorithm

With the development of synthetic aperture radar (SAR), it has been widely used in thenational defence and civil industry. Especially for battlefield reconnaissance, disaster forecastand resource mapping, SAR is playing an increasingly important role in modern society. SinceSAR is capable of observing earth surface day and night, all-weather, long range, highresolution, wide area, higher resolution and wider swath are required in updating SAR system.On the one hand, spaceborne SAR can be an effective way for high resolution and wide swathSAR imagery. However, due to the conventional constraint of minimum area on SAR antenna,high resolution in azimuth and wide swath in range are incompatible. This dissertation studiesthe methods breaking through the constraint based on multi-channel of single satellite anddistuibuted micro-satellites, respectively, and achieves SAR imagery with high azimuthresolution and wide range swath at mean while. In the study, real characteristic of satelliteorbit and earth rotation and3dimentional (3-D) baseline among micro-satellites are taken intoconsideration. What is more, the Doppler ambiguity removal with beamforming technique isapplied to realize SAR imaging under low pulse repetition frequency (PRF) sampling. On theother hand, airborne SAR has fewer limitations and is easy to be realized which can be usedto achieve higher resolution SAR imaging. However, due to the inevitable turbulence inaircraft flying, motion error is introduced into the echoes. Especially for miniaturizedunmanned aerial vehicle (UAV), motion error compensation is crucial for high quality SARimagery. Also in this dissertation, we proposed three motion compensation algorithms tocompensate the motion error accurately, efficiently, and in robust. All of these algorithms canachieve motion compensation in data-driven adaptive manner, even when the precision ofinertial navigation measurement is limited in the experiments. The main content of thisdissertation is summarized as follows:1. Study on high resolution wide swath SAR imagery using single satellitemulti-channel system. The multiple input and multiple output-Scan (MIMO-Scan)mode is designed and the related signal model is derived. By adopting multi-channelin azimuth and the beamforming technique, Doppler ambiguity can be removed toachieve high azimuth resolution under low PRF sampling. Besides, bandwidthsynthesis is utilized to obtain high range resolution at low cost of SARimplementation. SAR parameters are designed considering real application.2. Study on high resolution wide swath SAR imagery using distuibuted micro-satelliteconstellation. The micro-satellite in ideal linear constellation and ideal constellation with3-D base lines are analyzed. The SAR imagery algorithm combined withDoppler ambiguity removal is implemented under sub-aperture coordinates takingbaseline errors into consideration. The effects of satellite ellipse orbit and earthrotation is also considered. The method of coordinate equivalent transformation toovercome the geometric complexity from ellipse orbit and earth rotation is proposed.3. Study on motion compensation (MoCo) based on Doppler rate estimation for highresolution SAR imagery. The relationship between the precision of MoCo andresolution required is analyzed, which is confirmed by an UAV raw data. The echosignal model based on Doppler rate estimation is derived, and the phase error inquadratic is obtained, which corresponds to the acceleration of the motion error.4. Study on MoCo based on WPGA for high resolution SAR imagery. The error modelis established first, and then the PFA is applied for SAR image formation. The errormodel after PFA is derived accordingly, and the NsRCM and phase error, which areboth induced by the motion error, can be obtained quantitatively. By adoptingwavefront curvature correction method in the SAR image formation, wide swath canbe observed. The processing flow with two steps, the coarse and fine MoCo, isdesigned to solve both NsRCM and phase error completely. Both range-invariantand variant motion error are taken into consideration in the compensation.5. Study on MoCo based on entropy minimazation for high resolution SAR imagery.Convetional MoCo based on minimum entropy principle is analyzed first, includingminimum entropy range alignment and minimum entropy phase autofocusig. In theconsideration of coherence between the error in RCM and phase error, jointautofocsing that can realize range alignment and phase autofocus simultaneously isdesigned based on entropy minimization. According to the common source of RCMand phase error, we design a novel iterative strategy based on phase corrected andRCM corrected data alternately for each iteration. This can improve the efficiency ofconvergence. Besides, normalized polynomial fitting in high order case is adopted tosimplify the scale of the optimization problem.