Euclidean Distance Matrix-Based Rapid Fault Detection and Exclusion

Faulty signals from global navigation satellite systems (GNSSs) often lead to erroneous position estimates. A variety of fault detection and exclusion (FDE) methods have been proposed in prior research to both detect and exclude faulty measurements. This paper introduces a new technique for the FDE of GNSS measurements using Euclidean distance matrices. After a brief introduction to Euclidean distance matrices, both the detection and exclusion strategy is explained in detail. Euclidean distance matrix-based FDE is verified in two separate real-world data sets and proven to accurately detect and exclude GNSS faults on an average of 1.4-times faster than residual-based FDE and 70-times faster than solution separation FDE.

Automated summary

This paper introduces a new technique using Euclidean distance matrices for fault detection and exclusion of GNSS measurements. The technique is proven to be accurate and significantly faster than other existing methods.