GPS SISRE/URA Integrity Analysis for ARAIM

Abstract

The modernization of current GNSS services such as GPS, and the development and deployment of new constellations, such as Galileo, allow for new concepts as ARAIM to be proposed and raise interest among the aviation community. The ARAIM concept was proposed in the framework of the GNSS Evolutionary Architecture Study (GEAS) to define a seamless worldwide GNSS-based navigation alternative for various aircraft operations. The ARAIM architecture requires relatively simple ground infrastructure as the user algorithm can deal with part of the integrity assurance, but this user algorithm relies on a number of assumptions. These relevant assumptions are mainly related to the user algorithm input parameters, such as the probability of failure of a single satellite, and significant effort is being devoted to assess to which extent these assumptions are realistic. One of the main implications is that the satellite’s Signal in Space Range Error (SISRE) is assumed to be bounded by a Gaussian distribution with a given mean and variance to some probability level. In GPS, the nominal signal accuracy is provided by the navigation messages in the one-sigma parameter called User Range Accuracy (URA), which is aimed at representing a conservative overbound of the SISRE. This paper aims at providing an insight of the SISRE/URA behavior of GPS over several years of real GPS data, including a characterization of the SISRE associated to the different individual satellites and satellite blocks, a summary of the major events detected in the period of time, and estimates for the satellite faults probabilities.The strategy for this work is based an offline GPS data analysis that compares the ephemeris data broadcast by GPS satellites with the precise orbit data computed by IGS (International GNSS Service). IGS makes use of a global network of GNSS receivers to record the GPS navigation, and produces, among other products, the so-called ‘brcd’ files that contain the GPS broadcast ephemeris, and the precise GPS ephemeris and clocks computed by IGS, which are used as a reference, as the ‘true’ satellite’s positions and clocks. IGS has been chosen for this study because it provides an open and well known source of GNSS reference data. The methodology of the study has as a starting point the SISRE error computation directly from the ‘brdc’ and precise ephemeris and clocks provided by IGS. Then these preliminary results are screened for potential events, meaning unexpected high error values, which are then further investigated to look for potential inconsistencies in the ‘brdc’ input files. These files contain a consolidated version of the GPS navigation messages received by all stations in the IGS network over time, so we can cross check the ‘brdc’ file contents with the navigation files provided by each individual IGS reference station. Finally inconsistencies found in the ‘brdc’ are removed. Once the SISRE over the full period of the study is computed, a statistical analysis is performed, showing the distribution of the error and its relation with the broadcast URA. As a branch of the statistical analyses, the NANUs issued for GPS have been studied under the hypothesis of ARAIM users being able to receive and process the NANUs. The impact of this hypothesis on the probability of satellites’ faults is assessed and compared with the main analysis results. This study has been developed as part of ESA’s MULCOBA project.