UAV-Enabled Charging Path Planning for IoT Sensors With Adaptive Dormancy Using Reinforcement Learning

As wireless charging technology advances, autonomous aerial vehicles (UAVs) are emerging as a primary energy source for Internet of Things (IoT) sensing networks. However, the limited power capacity of UAVs poses significant challenges in ensuring timely and efficient energy replenishment for IoT devices. Traditional approaches rely on fully charging devices in each cycle, which can be inefficient. This article introduces a partial charging model aimed at maximizing system utility by accounting for the energy consumption patterns of sensing nodes and the UAV’s optimal hovering positions. The proposed system utility integrates key objectives: maximizing total data collection, minimizing the number of dead nodes, and minimizing the UAV flight distance. We propose a scheme recharge and adaptive scheduling for energy (RASE) to maximize the system utility. RASE scheme combines reinforcement learning with a sensor node dormancy strategy based on a matrix completion algorithm while also introducing an adaptive charging power strategy tailored to the data collection needs of individual sensing devices. Extensive simulations validate the effectiveness of the proposed algorithm, demonstrating significant improvements in system performance. The RASE scheme not only maximizes the effectiveness of the system but also greatly extends the lifetime of the sensing network.