An Efficient Docking-Point Deployment and Charging Access Coordination Method for Embodied-Enhanced UAV Networks

Abstract

As embodied intelligent agents, uncrewed aerial vehicles (UAVs) support low-altitude urban services, but their endurance is fundamentally constrained by limited onboard battery capacity. Existing solutions in dense urban environments incur high deployment costs, use coarse spatial layouts, and do not scale to large UAV fleets. We instead retrofit existing urban deployable infrastructure (UDI), such as traffic lights, street lamps, and communication base stations, as UAV docking points with charging capability. This UDI-based approach raises two coupled challenges: city-scale docking-point deployment over massive, spatially heterogeneous candidates, and coordinated multi-UAV access under queueing delays and residual-energy safety constraints. We jointly model docking queues, load, and energy consumption, and formulate a multi-objective optimization balancing energy consumption and load. To address these NP-hard deployment and scheduling subproblems, we propose a hierarchical UDI-based docking-point deployment algorithm (HUDD) that generates a scalable docking layout, and a charging access coordination algorithm based on convex relaxation and iterative rounding (CRIR) that coordinates energy-feasible, congestion-aware access for multiple UAVs on the obtained layout. Simulations on realistic urban datasets show that HUDD-CRIR outperforms baseline schemes in terms of energy consumption, response delay, queueing delay, and load distribution.