Research On Airborne SAR Imaging Based On Real Data

Synthetic Aperture Radar (SAR) possesses the advantages of all-time, all-weather, two-dimensional high-resolution, strong penetrating and the characteristics of signal processing to obtain image, it plays a very important role in military and civilian fields. Compared with spaceborne SAR system, Airborne SAR system is the primary objective of national development for reasons as follows: easy to implement, more flexible, real-time imaging, and generally the spaceborne SAR technology is verified through the airborne SAR.To combine the airborne SAR system development of a research institute of CASIC, the paper mainly studies airborne SAR imaging algorithms of RD and CS, Doppler parameter estimation and phase gradient autofocus (PGA).Firstly, the paper carries out theoretical study on RD and CS algorithm, point target simulation with RD algorithm are executed, in which range migration must be considered; In Doppler centriod estimation, spectrum fitting and correlation function (CDE) are studied and compared, according to the results of CDE, Doppler global estimation is implemented through N-M simplex method. And map drift (MD) and contrast optimization autofocus algorithm (COAA) are researched and contrasted in Doppler chirp rate estimation. In order to obtain the quick images of a large number of scenes, it designs two kinds of quick imaging algorithms of RD. Through the contrast of quick images, the chirp rate estimation with COAA gets a better result in case of stable platform and greater contrast.Secondly, the paper studies the platform motion error classification and compensation, analyzes the motion characteristics of vehicle platform according to the measured data of IMU. RD algorithm is used for point target simulation of motion compensation, real data imaging and motion compensation. However, the effects are not very well because of the instability of vehicle platform. It analyzes the motion characteristics of airborne platform, for the case of measured velocity error, the CS algorithm combining COAA is adopted. Comparison of the images show that the CS algorithm combined with COAA can effectively estimate the true speed.At last, In order to obtain the required resolution of practical engineering, for handling some defocus images, the paper researches on PGA algorithm. For images with corner reflectors and without corner reflectors, estimating the window length automatically and fixed initial window length then reduced in turn in proportion are analyzed and compared separately, processing results show that the former is more flexible when contrast is high, the latter is more stable on the whole.