A Multi-hop Data Routing Method Using UAV Swarms Based on Improved UCB for Power Emergency Communication

In the context of frequent natural disasters worldwide, we propose a multi-hop data routing method using unmanned aerial vehicle (UAV) swarms based on improved upper confidence bound (UCB) algorithm, aimed at solving the problem of efficient data backhaul in power emergency communication scenarios. First, a data routing model is constructed considering the data packet backhaul process in power emergency communication. Then, an optimization problem is formulated with the objective of minimizing the total delay for transmitting all data packets from ground repair personnel to the command center via UAV relays. Next, the original problem is transformed into a multi-armed bandit (MAB) problem, and a data routing optimization algorithm based on improved UCB is proposed. This algorithm adaptively adjusts the preference values for arms in real time based on data packet size, achieving a balance between exploration and exploitation and optimizing the data packet routing decision. Simulation results show that the proposed algorithm reduces total delay by $8.08 \%$ and $27.96 \%$ compared to two baselines.