Study Of High Resolution SAR Imaging Based On The Navigation Data

The Synthetic Aperture Radar(SAR) has been an important reaserch field of high resolution SAR imaging based on navigation data. In the process of Back Projection(BP) algorithm, the distances between the Antenna Phase Center(APC) and every pixel will be calculated accurately. By using the navigation data of the APC, the motion errors could be compensed. It shows the research value in the detection of high resolution SAR imaging.In this thesis, the fitting method of IMU data、the Streaming BP imaging algorithm for airborne SAR and the performance of the parallel programming will be analyzed. The high resolution SAR imaging result based on the measured data will be get finally, the main contents in this thesis are as follows:First, in order to meet the needs of SAR BP imaging’s process, the coordinate system’s conversion and the fitting method of IMU data of navigation in the IMU system has been studied. A piecewise of high-order polynomial fitting method has been implemented. Besides, the performance has been compared with some different fitting methods. Finally, the provement has been given that the piecewise of high-order polynomial fitting method can be applied to IMU navigation data processing, and the IMU data after fitting can realize the motion error compensation better, and the high resolution SAR images are obtained.Seccond, the BP algorithm has been studied and the process of BP algorithm is also given; This thesis proposed the Streaming BP imaging algorithm for the large amount of calculation for airborne SAR BP, and the SAR data in the imaging area will be processed with sliding window model. To realize the Streaming BP imaging algorithm, the GPU parallel programming platform has been studied which contains the memory management、cache structure design and scheduling policy of data management, etc; Finally, in Streaming BP method, the performance in the process of data processing has been studied, reaserching the running time of different processing module and optimization of the BP kernel module performance which costs longest time in the process.Finally, reaserch the motion error types of the aircraft and the impacts of SAR imaging based on the different motion errors. Calculating the carrier frequency in theory through the research of SAR echo’s time-frequency spectrum based on the measured data. The parameter estimation has been proved though the images’ contrast. Further, the flat-earth-removing of BP method for InSAR system has been proved in this thesis based on the actual Ku-band data, and the impact of attitude errors in APC has been studied in the end.