GNSS Signal Reliability Testing in Urban and Indoor Environments

The availability of two GNSS (Global Navigation Satellite System), GPS and Galileo, will offer in future new possibilities to provide integrity and reliability information to the user both at signal and user levels due to increased redundancy. User-level reliability monitoring schemes, namely Receiver Autonomous Integrity Monitoring (RAIM), consist of statistically testing least- squares residuals of the observations on an epoch-by- epoch basis aiming towards reliable navigation fault detection and exclusion (FDE). Classic RAIM and FDE techniques are based on only GPS characteristics, so in this paper, methods will be discussed also suitable for a combined GPS/Galileo system with the focus on personal location in degraded signal environments. This paper concentrates on analyzing different navigation quality and reliability assessment procedures based on testing the GNSS least-squares residuals on an epoch-by- epoch basis. The focus will be on reliability testing schemes for degraded GNSS signals in urban conditions in order to obtain an acceptable position estimate, and analyzing the urban GNSS navigation accuracy conditions. The reliability testing schemes for integrated GPS/Galileo to be discussed include applying a global test for detecting an inconsistent location situation, a local test for localizing and eliminating measurement errors recursively and, in addition, certain measurement subset testing. The proposed FDE schemes are examined with simulated GPS/Galileo data and real-life urban GPS tests. Furthermore, some external reliability measures, Mean Radial Spherical Error and Distance Root Mean Squared estimates approximating the effect measurement errors have on the accuracy will be analyzed. This paper will provide an insight into user-level integrity and reliability monitoring and FDE schemes eligible for a future GNSS system particularly for degraded signal environments, where the conventional assumption of normally distributed errors does not necessarily hold. The aim is to improve solution reliability and provide additional accuracy information to the user in terms of approximated position error estimates.