Study On Airborne High Resolution SAR Imaging And Location Techniques

Synthetic Aperture Radar(SAR) is a microwave active sensor, which can work all-weather, day/night and long range and can achieve fine-resolution images in two dimensions. As an important remote sensing tools are widely used in battlefield reconnaissance and civil surveying and mapping, etc. In recent years, with the continuous development of SAR technology, its application field has also been expanding. Now the SAR can not only achieve ultra-high resolution images of the designed field, but also give the precise position information of the targets in the field.Polar format algorithm(PFA) is widely used because of its high efficiency and easy to be combined with motion compensation, which is especially suitable for the large squint angle ultra-high imaging. Range Migration Algorithm(RMA) is commonly accepted to be an ideal solution of SAR focusing problem, which is particularly suitable for high resolution imaging of the large scenes. In this dissertation, both the PFA and RMA are studied in details and are used to dealwith the processing of SAR raw data which were collected by different airborne SAR system. In the engineering practice, the accuracy of inertial navigation system/inertial measurement unit(INS/IMU) is not high enough to get the accurate motion information, which would lead residual range migration in the ultra-high imaging. Then the point target response would be defocused in two dimensions. On the other hand, the geolocation accuracy would lose. In the high resolution wide swath imaging, the compensation of space-variant motion error is also problem. In this dissertation, the existing algorithm is further perfected and improved to solve the problems in the design and manufacture of airborne SAR systems.Chapter 1 introduces the research background and significance of this paper, and then reviews the development of SAR technology, analysis of high resolution SAR imaging and SAR geolocation technology research status in China and abroad. Finally this paper introduces the main research content of this article.In chapter 2, the airborne SAR high resolution imaging algorithm is studied. First, the SAR signal model is built, on basis of which the PFA and RMA is reviewed. In PFA, the range scale transforming and the space-variant post-processing for wavefront curvature correction are introduced. For the RMA, the extended algorithm with two-step motion compensation is reviewed. Finally, the two algorithms are respectively validated by processing the point target simulation data and the real data.In chapter 3, the residual range cell migration(RRCM) correct algorithm is studied. Firstly, the analytic expression of motion error in PFA is deduced, which reveals that how the residual range migration occurs. Then aperture image drift method is introduced into fitting the RRCM of low frequency. To overcome the high frequency RRCM, an autofocus method integrated with two dimensions inverse filtering is proposed. The new method can correct the RRCM effectively and estimate the phase error accurately.In chapter 4, the high resolution wide swath imaging algorithm of airborne SAR is studied. Firstly, the influence of space-variant motion errors to the imaging is formulated, and two methods are presented to compensate the space-variant motion errors. The first method is the ERMA based on sub-band along range direction, and the other one is based on numerical calculation. Lastly, an autofocus algorithm for the strip-map SAR is studied.In chapter 5, the airborne autonomous location technology is studied. Firstly the airborne SAR geolocation accuracy is discussed and the analytic term of geolocation errors are deduced. Secondly, the geolocation method based on Range Doppler(RD) model is introduced, and a new location method using imaging the area of the target repeatedly is proposed. Real data processing results demonstrate that the new method proposed can improve the location precision significantly. Lastly a location algorithm based on the least square method is analytically derived, which can improve the location precision further.