A High-Fidelity Wideband Signal Software Simulator for GNSS Antenna Arrays Accelerated by GPU

In order to reduce the impact of interference, multipath, spoofing, etc. on the performance of GNSS (Global Navigation Satellite System), array processing methods have attracted more and more attention in the past years. Developing, testing and evaluating array processing methods need a reliable, flexible platform. This paper proposes a high-fidelity array signals generating method, and it is implemented completely in software. The software calculates the relative delay of different antenna elements and generates narrowband or wideband GNSS/interference signals by the delay. Simulation results demonstrate that the array signals generated by the fractional filter are more capable for evaluating the performance of the array processing methods in actual environment than the array signal generated based on the narrowband assumption. Moreover, the software simulator is focused on the simulation of non-ideal factors in real test environments, such as antenna pattern, antenna mutual coupling, and non-ideal characteristics of the RF channel. Capability of the simulator to simulate antenna mutual coupling and non-ideal characteristics of the RF channel is demonstrated with several examples and comparisons in antenna pattern , amplitude and phase of the signal. A CUDA-enable GPU is utilized to solve the problem that the simulator requires a large amount of computing resources for the formation of IF data of each channel, the signal filtering process, and the simulation of the mutual coupling. The simulator fully utilizes the parallel computing capabilities of a CUDA-enable GPU and optimizes the algorithm details to improve the efficiency of the software simulator. Compared with traditional methods, the software simulator could provide more realistic narrowband and wideband signals in a reliable, flexible, and efficient way.