Performance analysis of CUDA based Galileo Signal Quality Monitoring

The aim of this work is to develop basic findings for a continuous Signal Quality Monitoring system based on a measurement campaign. Four Galileo satellites were repeatedly recorded and their metrics were analyzed. Due to the stable course, thresholds for the detection of threat models can be determined. These values were tested against simulated signals and the sensitivity of the detection was found to be satisfactory. Based on the convergence behavior of the data, a recommended measurement duration of 180-200 seconds can be recommended. Finally, the influence of the GPU and memory clock on the performance of predefined conditions close to the receiver was tested. The core clock of the GPU was identified as the bottleneck of the processing.