Application of reinforcement learning in UAV cluster task scheduling

Recently, unmanned aerial vehicle (UAV) clusters have been widely used in various applications due to its high flexibility, large coverage and reliable transmission efficiency. In order to achieve the collaboration of multiple UAV tasks within a UAV duster, we propose a task-scheduling algorithm based on reinforcement learning in this paper, which enables the UAV to adjust its task strategy automatically and dynamically using its calculation of task performance efficiency. As the UAV needs to perform real-time tasks while working in a dynamic environment without centralized control, it needs to learn tasks according to real time data. Reinforcement learning has the ability to carry out real-time learning and decision making based on the environment, which is an appropriate and feasible method for the task scheduling of UAV clusters. From this perspective, we discuss reinforcement learning that solves the channel allocation problem existing in UAV cluster task scheduling. Finally, this paper also discusses several research problems that may be faced by the further application of UAV cluster task scheduling. (C) 2018 Elsevier B.V. All rights reserved.