Analysis of the Usability of an SBAS Correction Solution for Smart Mobile Devices

Abstract

Mobile devices are now powerful tools for an increasing number of applications in the field of geomatics. In this study the application of SBAS (Satellite Based Augmentation System) correction data on measurements of smart mobile devices is investigated to improve the performance of positioning capabilities in terms of accuracy and precision of the solution. For that purpose, a correction model is developed employing SBAS correction data on both single- and dual-frequency multi-constellation Android-based smart mobile devices. A comparison of the smart mobile devices’ results was made to two geodetic grade GNSS receivers. As the observations in this study with these devices were carried out in Sri Lanka, the Indian GAGAN (GPS Aided Geo Augmented Navigation) SBAS is used for the correction model derivation. It must be noted, however, that the developed approach is capable of working with any SBAS worldwide. Before the SBAS correction application on the observations, the achieved Root Mean Square Error (RMSE) was around 75.2 and 33.3 m on average on single and dual-frequency smartphones, respectively. However, after applying the SBAS correction, all devices improved performance. The RMSE values improved for all devices, with the Stonex 900A geodetic receiver having the lowest value (0.42 m), followed by the Leica GS15 (0.49 m), the Android Stonex S70 GIS tablet (1.09 m), the Xiaomi Redmi Note 8 single-frequency smartphone (1.35 m), and the Google Pixel 4 dual-frequency mobile device (1.62 m). Similarly, the standard deviation (STD) values also improved, with the Stonex 900A and Leica GS15 having the lowest values. The SBAS correction also reduced the size of the confidence error ellipses for all smart devices. After SBAS correction, even the single-frequency carrier-phase smartphone Redmi Note 8 showed a significant reduction in SD and RMSE values (98% improvement), which surpassed the performance of the Pixel 4 in standalone mode. This can be attributed to SBAS correction relying solely on the L1 frequency signal and correction (such as in the GAGAN augmentation), independent of the L2 or L5 GPS frequency. Additionally, differences in hardware and software between devices can impact GNSS performance, including antenna design, processing algorithms, and signal filtering techniques, all influencing GNSS accuracy and reliability.