Second Generation SBAS – Performance Analysis and Bridging Positioning and Integrity Monitoring during SBAS Outages in the Urban Environment

Abstract

Improved SBAS is expected to be a primary positioning method for many land applications such as Intelligent Transport Systems (ITS). In 2017, a second-generation SBAS test-bed was initiated in Australia and New-Zealand in preparation for building an operational system. In addition to the traditional L1 legacy SBAS signals, the Australian/NZ SBAS provides dual-frequency multi-constellation (DFMC) SBAS signals and SBAS-based PPP service for GPS and Galileo observations. This article addresses the use of SBAS in land applications from three important aspects. Firstly, the performance of the different SBAS solution methods is assessed in two main applications, transportation and mining. Tests were conducted under various environments, including open sky, low- and high-density urban, and mining. Performance analysis showed that SBAS solutions provide better positioning than single point positioning, and the DFMC SBAS solutions have better precision than L1 SBAS solutions. Furthermore, SBAS-based PPP solutions delivered a few-dm accuracy, which is a bit worse than traditional PPP with corrections received via the Internet. It is shown that SBAS performance is strongly dependent on the application environment. Secondly, tests show that SBAS outages may occur in urban areas, thus, the paper proposed prediction of the SBAS orbit and clock corrections as time series to enable positioning. The orbits are predicted using Holt-Winters’ method and clock corrections were predicted using a second order polynomial with sinusoidal components. A position error due to prediction grow with time, but was less than 0.26m after 15 minutes of prediction. Finally, the impact of SBAS outages on computation of the protection levels (PL) needed for integrity monitoring of the new DFMC SBAS is illustrated. It is shown that Horizontal-PLs can reach more than 10 m, which is not suitable for ITS when the previous SBAS parameters before the outage were used, and thus methods are proposed for future research.