A Software-Defined Real-Time GNSS Signal Quality Monitoring Receiver using GPU(Graphic Processing Unit)

In this paper, we implemented the software-defined GNSS signal quality monitoring receiver using GPU. Two channel 50MSps-sampled intermediate frequency (IF) signal are acquired and tracked in real-time. This monitor shows results such as correlation shapes of each satellite signal, integrated waveform, positioning results, and etc. In the case of using a NVIDIA GPU(K20), The implemented receiver can track the 36 channels. Each channel has the 64-tap code delay (variable code delay) outputs. For the real-time implementation of the massive correlators, a modified reduction technique was proposed and strategic use of memory such as a register, a constant memory, and a global memory was considered. Additionally, a spectral characteristic of the IF signal can be monitored. The developed system is consisted of two RF front-end device, two 16-bit digitizer, three NVIDIA GPU, and a workstation. The system can process the dual-frequency signals in real-time.