Design and implementation of aerial communication using directional antennas: Learning control in unknown communication environments

Abstract

To support an increasing number of commercial multi-unmanned aerial vehicle (UAV) applications, robust UAV-to-UAV communication is critical. A promising solution is aerial communication using directional antennas (ACDA), with features like long communication distance, low power consumption, broad bandwidth, and interference rejection. Nevertheless, ACDA requires the automatic alignment of directional antennas, which is not easy to achieve considering the imperfect communication environment unknown in advance and the limited sensing devices onboard due to the constrained payloads and power sources. In this study, the authors design and implement a new ACDA system, including the platform, communication, computing, control, middleware, and interface components. Practical implementation issues for the emergency response application are also considered. The ACDA system features a communication and control co-design, where the communication quality indicator, received signal strength indicator, serves as the goal function for antenna alignment. The solution also features a reinforcement learning-based directional antenna control algorithm that learns the unknown communication environment models. The performance of the ACDA system is verified using simulation studies, field tests, and disaster drills.