Trusted Multisource Fusion Navigation for UAV Under GNSS Interference and Spoofing Attacks
Chen Meng, Qinglei Hu, Shuzhi Sam Ge, Dongyu Li · IEEE/ASME Transactions on Mechatronics · 2025
Trusted navigation is paramount for the safe operation and mission success of unmanned aerial vehicles (UAVs) in complex environments. However, individual navigation systems possess inherent limitations. For example, global navigation satellite systems (GNSSs) are vulnerable to signal denial, spoofing, and jamming attacks, whereas inertial navigation systems suffer from cumulative error growth that ultimately degrades navigation accuracy. To overcome these challenges, this article proposes a novel multisource fusion navigation framework tailored for UAVs operating under adverse conditions. First, we introduce a mechanical antenna (MA)-based positioning method as a complementary solution to GNSS unavailable. The MA, mounted on the UAV, transmits low-frequency magnetic field signals captured by ground-based magnetic sensors; subsequently, a particle swarm optimization algorithm is employed for MA positioning. Recognizing that near-field magnetic interference adversely affects MA accuracy, we further integrate an anomaly detection mechanism with dynamic adaptive fusion and federated filtering algorithms, respectively, to isolate data from compromised sensors. Comprehensive experimental evaluations, including UAV flights under GNSS spoofing attacks, demonstrate that the proposed fusion approach not only effectively detects various sensor attacks (achieving average accuracy and precision of 99.7% and 98.7%, respectively), but also maintains trusted navigation in GNSS-challenged and magnetically interfered environments.