An attempt to model the accuracy of GNSS Positioning from GIPSY-X PPP-AR Referring to  JPL’s 2019 Experiment

Abstract

Multi-GNSS techniques are continuously improving themselves and strengthening their infrastructures. In response, companies producing GNSS hardware and software are enhancing their products with refined techniques, leading to improved positioning precision. In line with these advancements, efforts are being made to enhance algorithms which predict the accuracy of GNSS positioning. Following these improvements, users who conduct field works or optimize networks would want to know about the improved accuracy of GNSS positioning. After GPS, the new members of the GNSS are continually improving their infrastructure. To achieve the desired coordination in terms of a common processing ground among these techniques remains a challenge. This is particularly crucial for organizations developing GNSS software. For instance, NASA JPL has been making efforts to produce a PPP-AR solution. The transition from GIPSY OASIS II to GIPSY-X allowed for the processing of combined GPS, GLONASS, and GALILEO data in a JPL experiment in 2019. However, today, there is yet no AR infrastructure for the combined solution of all techniques beyond the float solution. In this study, we investigated whether JPL's 2019 AR solution is still applicable to current research. Initial attempts yielded promising results as we conducted experiments with 50 points from the IGS MGEX network's shared data. By comparing the differences between the 2019 PPP-AR solution and the 2019 and 2023 float solutions, we found that the discrepancies are not significantly large in today's GNSS accuracy. We concluded that an accuracy model generated from the 2019 AR solution could provide a satisfactory estimate for today's users. Initial findings indicate that GPS performs well with the combination of three techniques, while the synergy of techniques significantly contributes to the improvement of the vertical positioning. The accuracy produced by the combination is witnessed to be dependent on the observation period, emphasizing the need for attention from those conducting campaign measurements in this context.