A Reinforcement Learning-Based Resource Allocation Method for UAV Swarms

To address the issue of low task completion efficiency caused by dynamic changes in network resources during UAV swarm missions in remote and unknown environments, this study models the UAV resource allocation problem as a system throughput maximization problem. To enhance communication efficiency in UAV communication systems, the problem is formulated as a partially observable Markov decision process, and a high-efficiency resource allocation mechanism based on deep reinforcement learning is proposed. Specifically, a proximal policy optimization (PPO) gradient mechanism is introduced to achieve optimal resource allocation aimed at maximizing average throughput. Simulation results validate the effectiveness of the proposed algorithm.