Efficient analysis of GPS antenna onboard self-driving car in practical environments with GPU accelerated MoM-PO method

The modern graphics processing unit (GPU) programming technique is applied to speed up the simulation of the performance of GPS antenna installed on a self-driving car in practical environments with the method of moments (MoM) and physical optics (PO) hybrid technique. To take good advantage of GPU acceleration, the MoM part, ray tracing, PO current calculation, and post processing are all executed on GPU by developing a set of GPU code with CUDA. Numerical results show excellent accuracy and efficiency improvement.