GNSS Measurement Fault Detection for Urban UAV Swarms Based on Dual-Mode Parity Vector Assisted by Relative Ranging

Acquiring precise and continuous positioning measurements is critical for urban unmanned aerial vehicle (UAV) swarms, yet urban canyon environments frequently induce global navigation satellite system (GNSS) measurement faults. These anomalies not only degrade individual UAV performance but also risk swarm-wide positioning drift, severely threatening mission safety. While existing fault detection and exclusion (FDE) methods face a critical dilemma - solution separation techniques are robust but computationally inefficient for multifaults, whereas conventional parity vector methods are efficient but struggle to distinguish multiple simultaneous faults and are insensitive to gradual faults - a comprehensive monitoring solution is urgently required. This article proposes a novel dual-mode parity vector with relative ranging (DMPV-RR) fault detection method to overcome these limitations. The proposed approach enhances the traditional parity vector mechanism by integrating three key innovations: 1) a subspace projection metric that achieves high-accuracy multifault identification without excessive computational overhead; 2) the incorporation of inter-UAV relative ranging measurements to establish network consistency constraints, thereby significantly accelerating the detection of gradual faults; and 3) an adaptive thresholding mechanism that dynamically adjusts fault decision criteria based on real-time satellite geometry and observation quality. The effectiveness of the DMPV-RR method was validated through a comprehensive three-stage methodology: software simulation, hardware-in-the-loop (HIL) simulation, and real-world vehicle experiments. Results demonstrate that the proposed method achieves a 98.5 % detection accuracy for small-magnitude abrupt faults and, critically, reduces the detection latency for gradual faults by up to 9 s in highly dynamic practical scenarios. The DMPV-RR method provides a validated, highly reliable fault monitoring solution for safeguarding the reliability of UAV swarm integrated navigation systems in challenging urban environments.