A Research On Real-Time UWB-SAR Imaging Algorithm Based On Subaperture

Ultra WideBand Synthetic Aperture Radar (UWB-SAR) has a broad military application because it can produce fine image of concealed targets with the capability of penetrating the foliage and the earth surface. Real-Time UWB-SAR imaging is of great importance in target detection and recognition. Real-Time UWB-SAR imaging is computationally intensive, and therefore the efficient imaging algorithm is the pivotal and foundation.This dissertation first.analyzes the properties of UWB-SAR imaging and investigates its impact on the real-time implementation. Based on the simplified real-time imaging system model, the complexity analysis of UWB-SAR imaging is carried out according to the Range-Doppler imaging algorithm. After the comparison of various imaging algorithm, several of them are selected for further studying. The BackProjection (BP) algorithm is presented and the simulation shows that efficiency is sharply decreased by the interpolation. Therefore BP algorithm denies its application in wide swath UWB-SAR imaging. In the end, the Frequency Scaling and Chirp Scaling Algorithm are chosen for further study because of their computational efficiency.The FSA is originally developed to process spotlight SAR with dechirped echoes. After a detailed interpretation of frequency scaling principles and a discussion of the algorithm, we propose an improved FSA for UWB-SAR. The algorithm improved the performance of its standard counterpart considerably, while the computational complexity is almost unchanged. Besides, the subaperture structure is introduced into the FSA .As the UWB-SAR resembling the high squint SAR in image formation, the NCS (Nonlinear CS) is studied. After the analysis of the nonlinear filtering principles, a modified NCS algorithm is presented, which yields well focused image and greatly improves the swath width. Further more, a subaperture structure is applied to the algorithm, and this could make use of multiprocessor for parallel computation and reduce the memory requirement as well. The subaperture NCS is an appropriate algorithm for UWB-SAR real-time imaging. At last, we bring forward a scheme for the implementation of the subaperture NCS algorithm with multiprocessor.All the algorithms in this thesis are verified by computer simulation and UWB-SAR raw data processing.