Evaluating The Performance of Open Source GNSS Online Post-Processing Services

Abstract

Various organisations have recently developed online processing services for GNSS data. However, due to limited information about their performance and accuracy, users struggle to make informed decisions. The objectives of this study are to assess the accuracy of the positioning solutions provided by the three online GNSS processing services (OPUS, AUSPOS, and CSRS-PPP) and determine the optimum observation times needed to achieve high accuracy in GNSS post-processing services. The study was conducted at UiTM Shah Alam, Selangor, with field observations performed at four selected control points using the static observation technique. The coordinates of control points were estimated using TBC software and online processing services, and the differences between the coordinates from these services and TBC software were computed. The results indicate that AUSPOS outperforms other services in terms of accuracy across different observation periods (4 hours, 3 hours, and 2 hours). For the 4-hour observation, the RMSE values are ±0.020 m for northing, ±0.022 m for easting, and ±0.028 m for ellipsoidal height. In the 3-hour observation, AUSPOS also shows the smallest RMSE values of ±0.028 m in northing, ±0.023 m in easting, and ±0.034 m in ellipsoidal height. Similarly, in the 2-hour observation, AUSPOS maintains the best performance with RMSE values of ±0.067 m for northing, ±0.073 m for easting, and ±0.082 m for ellipsoidal height. In conclusion, the study demonstrates that among the three online GNSS processing services evaluated, AUSPOS consistently delivers the highest accuracy across different observation durations, particularly in the 4-hour observation. While CSRS-PPP follows closely behind, OPUS shows the least accuracy, especially for shorter observation periods.