IEEE Access (Jan 2025)
Prediction of Ultra-High-Speed Spots Using RTK-GNSS Sensor Fusion for UAV-to-UAV mmWave/THz Communications
Abstract
mmWave/THz communications depend on highly focused and directive narrow beams. Accurate prediction of the beam’s spatial location and attitude in space, as well as the time it takes to reach its ultra-high-speed coverage area, referred to as the ultra-spot, is crucial for uncrewed aerial vehicles (UAVs) to adjust their flight direction and approach velocity. This adjustment increases the likelihood of successful communication between UAVs. This paper introduces a novel approach for detecting these ultra-spots using real-time kinematic (RTK)-GNSS and inertial measurement unit (IMU) sensor fusion positioning powered by an extended Kalman filter (RTK-GNSS-EKF). To achieve this, we implemented a mechanism that exchanges six degrees of freedom (6-DOF) information of positions among UAVs via a 920MHz wireless communication link. Additionally, we propose an algorithm that accurately estimates the time and distance from the in-flight UAV to the ultra-spot. For the first time, this work investigates the real-world 6-DOF fluctuations in position, velocity, and attitude experienced by an in-flight UAV due to wind, and analyzes the impact of these fluctuations on the ultra-spot prediction issue. Additionally, we analyze scenarios where the ultra-spot alters its attitude by actively changing the antenna angle, assessing the consequent effects on the volume of data transmitted and received at the ultra-spot. We demonstrate the effectiveness of the proposed method by simulation and verification with actual UAV-to-UAV and UAV-to-ground-station field experiments. Experimental results indicate an average ultra-spot detection accuracy of 172ms in time and 32.7cm in distance, with measurements taken 1s before the UAV’s actual approach to the ultra-spot. These findings confirm the feasibility of the proposed method for detecting mobile ultra-spots in UAV-to-UAV mmWave communication.
Keywords