Centralised cooperative UAV navigation in GPS-denied environments

Global navigation satellite systems (GNSS) are critical for both military and civil aviation, yet their susceptibility to jamming and spoofing poses significant threats to reliable navigation. This paper explores and quantitatively evaluates a centralised error-state Kalman filter (EKF) approach using range-only inter-UAV measurements for cooperative navigation of unmanned aerial vehicle (UAV) fleets operating in GPS-denied environments. The centralised ES-EKF leverages inter-UAV correlations by aggregating data from all UAVs in a central unit, enabling accurate state estimation even when multiple UAVs lose GPS access. Through simulations involving six UAVs with systematically varied GPS availability (from three down to a single GPS-enabled UAV), we demonstrate that the centralised ES-EKF significantly reduces navigation uncertainty and maintains mission integrity by utilising collaborative measurements. Scenarios where up to five UAVs experience GPS jamming illustrate the algorithm’s capability to sustain accurate positioning through cooperative data fusion, provided at least one UAV retains GPS functionality. This analysis provides clear performance benchmarks for such a configuration. While centralised ES-EKF offers superior performance in capturing inter-agent correlations, it also presents challenges related to computational load, scalability, and communication reliability in large networks. Therefore, this study established a baseline for future decentralised localisation works.