无线电通信技术

移动边缘计算(Mobile Edge Computing, MEC)技术在灾情救援、森林火警预警等对低延迟和资源稳定性要求苛刻的场景中应用日益广泛，然而地面基础设施匮乏常限制其效能。无人机(Unmanned Aerial Vehicle, UAV)凭借部署灵活性和高机动性，成为解决此问题的理想平台。创新地提出了一种均衡多UAV覆盖路径规划(Balanced Multi-UAV Coverage Path Planning, BmUCPP)方法，结合覆盖路径生成(Spanning Tree Coverage, STC)与最小生成树(Minimum Spanning Tree, MST)算法，重点解决多UAV协同作业中的负载失衡问题。针对边缘计算模型的多目标优化挑战，开发了改进的人工蜂群(Improved Artificial Bee Colony, IABC)-遗传算法(Genetic Algorithm, GA)的混合优化算法——IABC-GA,以最小化关键目标并保障MEC服务质量。测试表明，IABC-GA在寻优能力、收敛速度和稳定性上优势显著。为应对野外或灾区的实际需求，考虑UAV的通信、计算、续航限制，环境通信质量和地面用户设备(User Equipment, UE)能力，建立了一个动态UAV辅助MEC模型，旨在最小化UE与UAV的平均加权能效(结合能耗与时延)。通过深度结合所提BmUCPP与任务调度算法，多维度仿真证明该协同方案能有效降低动态UAV辅助边缘卸载的总体代价。

Mobile Edge Computing(MEC) technology is increasingly employed in scenarios demanding stringent low latency and resource stability, such as disaster rescue and forest fire warning. However, the scarcity of ground infrastructure often restricts its effectiveness. Unmanned Aerial Vehicles(UAV), leveraging their deployment flexibility and high mobility, serve as an ideal platform to address this challenge. This study innovatively proposes a Balanced Multi-UAV Coverage Path Planning(BmUCPP)method, combining the Spanning Tree Coverage(STC) algorithm with the Minimum Spanning Tree(MST) algorithm, with a primary focus on resolving the load imbalance problem in multi-UAV cooperative operations.To tackle the multi-objective optimization challenges inherent in edge computing models, an Improved Artificial Bee Colony(IABC)-Genetic Algorithm(GA)—IABC-GA is developed. This enhanced algorithm efficiently minimizes key objectives while ensuring MEC service quality. Evaluation results demonstrate that the IABC-GA exhibits distinct advantages in optimization capability, convergence speed, and stability.Aiming to meet practical requirements of field or disaster-stricken environments, a dynamic UAV-assisted MEC model is established. This model comprehensively considers UAV communication, computing, and endurance constraints, environmental communication quality, and capabilities of ground User Equipment(UE). The core objective is to minimize the average weighted energy efficiency for both UE and UAVs(integrating energy consumption and latency). By deeply integrating the proposed BmUCPP path planning algorithm with task scheduling algorithms, multi-dimensional simulations are conducted. The results substantiate that this collaborative strategy effectively reduces the overall cost of dynamic UAV-assisted edge offloading services.