Range-Only EKF-Based Relative Navigation for UAV Swarms in GPS-Denied Zones

Achieving successful collaboration among a swarm of Unmanned Aerial Vehicles (UAVs) requires an accurate relative localization system. UAVs operating autonomously using GPS-aided inertial navigation eventually require an alternative navigation aid during GPS outages. Without an alternative, these UAVs must rely on dead-reckoning navigation based on data from the Inertial Measurement Unit (IMU) alone, which can cause the location estimates to accumulate and the navigation solution to drift. This paper presents a range-aided navigation system, where an Extended Kalman Filter (EKF) combines IMU data with range observations from onboard sensors that measure the distance between nodes in the swarm. This method, typically called range-only EKF navigation, mitigates the growth of position, navigation, and timing (PNT) errors during periods of GPS signal loss. To demonstrate the effectiveness of this approach, a combination of software simulations and hardware experiments were conducted. These involved a swarm of six drones in various scenarios, including circular, square, and blue-angels flight trajectories. The trajectories were designed to minimize significant geometric dilution of precision (GDOP). Statistical analysis of the simulation results indicates that this approach improves the estimates compared with dead-reckoning alone, thereby reducing the growth of velocity and location errors when GPS is intermittently unavailable.