Design Of Missile-Borne SAR Simulation Processing Software Based On GPU

Synthetic Aperture Radar(SAR) is a kind of high-resolution imaging radar, which can be used to obtain high resolution radar images under adverse weather conditions. It is of great value in military.Study of SAR imaging method often requires the specific SAR original echo signal, which is difficult for researchers to obtain by actual radars. Echo simulation needs to calculate the echo for each scattering point of the target scene, which has a large amount of computation, and the time of echo generation is long. This has resulted in the difficulty of acquiring suitable echo data. In addition, the SAR echo data can also evaluate the correctness of the system designed, and help improve the performance parameters of the imaging system. CUDA is proposed by NIVIDIA for the general purpose of high performance computing. It is one kind of programming platform that greatly promoted the development of GPU in the field of high performance computing. Now, therefore, CUDA is widely used in the field of scientific computing. The process of SAR echo simulation also has a large degree of parallelism, which is suitable for the parallel implementation on GPU.With the development of Synthetic Aperture Radar imaging technology, combining the missile-borne SAR with the precision guidance technology has become a hot topic on research in recent years. In order to get more information of the target, it is necessary for the missile-borne SAR to get the high resolution, so as to make the shape and the details of the target more clearly, and thus it can greatly improve the ability of recognizing and attacking targets; In actual military application, missile-borne SAR usually need to attack after the investigation. In order to ensure that the missile has enough time to make a turn, SAR generally works in the highly squint mode. Therefore, it is important to study the Highly Squint missile-borne SAR imaging. At present, most of the parallel SAR imaging processing systems are based on the DSP and FPGA, which leads to the complexity of the systems, and the high cost of the hardware.The missile-borne SAR simulation software designed in this paper lets the missile borne SAR imaging processing more convenient and faster. Lab VIEW is used to design the UI interface that is easy to interact with and provides the input for parameters and data for echo simulation program and imaging processing program. In order to meet the requirements of real-time imaging and reduce the cost of the system, in this paper CUDA C language is used to realize the GPU programs for both the echo simulation and the missile-borne SAR imaging processing. This software not only ensures the correctness of the echo and the image result, but also achieves the effect of accelerating processing data.