Energy-Efficient UAVs and BSs Management in Distributed Edge Intelligence Empowered IoV Networks

The internet of vehicles (IoV) is playing a pivotal role in advancing intelligent transportation systems. Deploying the unmanned aerial vehicles (UAVs) as edge nodes in IoV networks has emerged as a promising solution to enhance the communication coverage and energy efficiency (EE). However, the existing UAV deployment and resource allocation strategies often lack the necessary intelligence and adaptability to respond to the dynamic traffic conditions. To address these challenges, we leverage machine learning (ML) technology to optimize EE by jointly optimizing small base station (SBS) dormancy and UAVs' three-dimensional (3D) positioning-a problem recognized as NP-hard. To tackle this problem, we propose an energy-efficient multi-UAV 3D deployment with SBS dormancy (MUD-SBS) algorithm, which decomposes the problem into two manageable phased issues. First, a dormant strategy based on the base station centrality (BSC) metric is developed to switch SBSs to a dormant state during low traffic periods. Second, the horizontal positions of UAVs are optimized using the k-means algorithm, followed by determining the optimal UAV heights via the genetic algorithm (GA). Extensive simulations validate the proposed algorithm can achieve a 41% improvement in EE and a 15% increase in communication coverage rate compared to the existing strategies. These results not only highlight the effectiveness of the proposed solution but also underscore its relevance in enhancing the performance and sustainability of IoV networks, paving the way for more intelligent and responsive transportation systems.