GPU-accelerated circular SAR echo data simulation of large scenes

摘要

In circular synthetic aperture radar (CSAR), echo data simulation can provide ideal echo data for the processing of imaging and performance analysis. However, when the target scene is large, the amount of echo data is also large according to Nyquist sampling theorem. If the traditional serial computing way based on CPU is used, then the execution time is long. In this paper, we adopt graphic processing unit (GPU) to improve the computing speed. This GPU-accelerated way can be simply described as the process that CSAR echo data with carrier information is generated on GPU, and the subsequent removal of carrier is implemented on CPU. Simulation results show the GPU- accelerated way is faster than the CPU-based way when the target scene is large.