Statistical Analysis of User Range Errors on BDS for ARAIM

Abstract

Signal-in-space errors (SISE) is mainly composed of broadcast ephemeris and clock error by comparing broadcast ephemeris and clock with the precise ones. SIS user range errors (UREs) are usually assumed as a zero-mean normal distribution, which needs to be overbounded by user range accuracy (URA). This article analyzes the BDS historical data in 2016 to determine whether the SIS UREs follow a zero-mean normal distribution with the standard deviation represented by URA and the impact of SIS UREs/URA to ARAIM. For the analysis of historical data, we first need to filter out the abnormal behavior in the ephemeris to determine the valid data and remove the difference of the time system between the broadcast ephemeris and the precise ephemerides. From the long- term statistics of SIS UREs, clock errors and orbit errors, the result shows that the SIS UREs is mainly determined by the satellite's clock performance. The SIS UREs are characterized with respect to mean, standard deviation, spatial correlation and normality. The results show that the mean of some satellites in BDS is not zero, the standard deviations of Medium Earth Orbit (MEO) satellites in BDS are less than 2m, the standard deviations of some satellites in Inclined Geosynchronous Orbit (IGSO) satellites are greater than 2m, and the standard deviations of all Geosynchronous Orbit (GEO) satellites are larger than 2m. Radial errors and clock errors in BDS are relatively less relevant, radial errors and along-track errors, along-track errors and clock errors are relatively strongly correlated. The URA value in Integrity Support Message (ISM) of an existing ARAIM simulation is usually 1m which is different from the statistical performance of BDS, and it is somewhat optimistic. In this paper, it is more appropriate to analyze the BDS SIS URA value of 2.8m. The URA value obtained from the historical BDS performance analysis was substituted into the ARAIM simulation (GPS + 35 BDS satellites worldwide). The results show that Horizontal Protection Level (HPL) and Vertical Protection Level (VPL) increase when compared with URA taking 1m, the HPL is still less than the Horizontal Alarm Level (HAL), the VPL will be greater than the alarm value, and BDS ARAIM availability decreases.