Characterizing the Android and iOS Internal Positioning Solution to Detect Wrong-lane Drivers

Christian A. Lichtenberger, Mohamed Bochkati, Thomas Pany

Abstract:	For location based services (LBS), especially when it comes to safety-critical applications, a both accurate and high-integrity position solution is essential. As part of the Ghosthunter project, in which wrong-way drivers are to be detected reliably on German highways, the accuracy of the positioning solution must be better than one lane width, i.e. less than 3 m, but the integrity of the position must also be guaranteed in order to avoid false alarms. So far, only Android smartphones have been examined in terms of the two performance parameters (accuracy and integrity) as part of this project. Since Android dominates the market for mobile devices with a share of just under 70% and iOS with a share of just under 30%, and since new application areas for precise and reliable positioning such as augmented reality (AR) or virtual reality (VR) and other safety critical applications are constantly growing, this work will examine the internal positioning solutions of both operating systems. Different iOS devices (iPhones, iPads) and some Android smartphones (including Xiaomi Mi 8) are available as test platforms. In direct comparison, the Android Location application programming interface (API) has also offered the option of directly acquiring the GNSS raw data and implementing one’s own positioning algorithms accordingly since 2016 (Banville and Diggelen, 2016), but the focus here will be on the various internal positioning modes to compare them with the internal positioning modes of iOS. Currently iOS does not offer the option of accessing the global navigation satellite system (GNSS) raw data. The Core Location API only provides positions and estimated accuracy. For iOS applications that require an exact knowledge of the position performance, this data must be used. To what extent the positions from the Core Location API fulfill integrity requirements for various applications has not been investigated yet in depth. For automotive applications, a measurement vehicle with geodetic GNSS equipment, namely Trimble NetR9 GNSS receiver with Trimble Zephyr geodetic GNSS antenna is available, which provides highly precise reference trajectories even in challenging environments. This will be used as ground-truth (GT) to analyze the positioning performance of the two operating systems under investigation. By comparing the trajectories in the kinematic case or the positions in the static case with the respective reference solution, the positioning performances (deviations, gaps, etc.) can be examined. By comparing the internally estimated accuracy and the true deviation from the reference solutions, statements can be made about the integrity of the positioning solution. On the one hand, a general statement shall be made, and on the other hand, different scenarios (shading by vegetation, urban environment, perfect open sky conditions) shall be considered separately.
Published in:	Proceedings of the 35th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2022) September 19 - 23, 2022 Hyatt Regency Denver Denver, Colorado
Pages:	1062 - 1073
Cite this article:	Lichtenberger, Christian A., Bochkati, Mohamed, Pany, Thomas, "Characterizing the Android and iOS Internal Positioning Solution to Detect Wrong-lane Drivers," Proceedings of the 35th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2022), Denver, Colorado, September 2022, pp. 1062-1073. https://doi.org/10.33012/2022.18415
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In