Energy Efficient Task Caching and Offloading in UAV-Enabled Crowd Management
Gaoxiang Wu, Qiang Liu, Jinfeng Xu, Yiming Miao, Matevž Pustišek
Abstract
Unmanned aerial vehicle(UAV)-enabled mobile edge computing(MEC) networks provide ubiquitous communication and computing capacity for mobile users compared with terrestrial networks. In crowd management, the UAV base station(UAV-BS) collect computation tasks from the Internet of Things (IoT) devices and process tasks with terrestrial MEC networks cooperatively. However, energy efficiency(EE) and user mobility are the bottlenecks of UAV performance. Therefore, it is crucial to maximizing the energy efficiency(EE) of UAVs. In this paper, we propose an energy-efficient UAV-enabled MEC network composed of IoT devices, the UAV-BS, edge cloud, and the data center, and propose a Green-UAV-CoCaCo algorithm to jointly optimize communications, caching, and computation for EE of UAV. Specifically, we design a UAV trajectory model based on a greedy algorithm to predict the user’s coordinates and choose the proper edge server for task offloading. Then, the UAV-CoCaCo algorithm is proposed to maximize the EE of the task caching and offloading. Simulation results demonstrate the effectiveness of the proposed algorithm.