An Inter-Frequency Cross-Validation Approach for Pseudo-Range Fault Detection in GNSS Relative Positioning

For Global Navigation Satellite System (GNSS) relative positioning, faulty pseudorange measurements may lead to over-bounded relative positioning errors, which entails high-performance fault detection and exclusion (FDE). This paper proposes an effective fault detection and exclusion method for pseudorange-based GNSS relative positioning utilizing the technique of the inter-frequency cross-validation (IFCV). Multi-frequency differenced pseudorange measurements are utilized to establish multiple inter-frequency test statistics for efficient detection of multiple outliers; the conservative strategy is adopted to exclude multiple faults for robust position determination. Compared with conventional ARAIM (Advanced Receiver Autonomous Integrity Monitoring) method, the experimental results indicate that the proposed IFCV method exhibits lower false alarm rates (0.03% vs. 1.88%) and missed detection rates (0% vs. 1.02%). By artificially injecting random faults into GNSS measurements, conventional differential pseudorange-based method shows a significant decrease in positioning accuracy by 354%, while both IFCV and ARAIM methods improve positioning accuracy by 78% and 55%, respectively. Apart from advantages in accuracy over ARAIM method, the proposed IFCV demonstrates a computational efficiency improvement of 10⁴ over ARAIM.