计算机科学
调度(生产过程)
无人机
分布式计算
能源消耗
实时计算
边缘计算
高效能源利用
移动边缘计算
软件部署
云计算
作业车间调度
动态优先级调度
无线
作者
Yangang Wang,Hai Wang,Xianglin Wei
出处
期刊:International Conference on Wireless Communications and Signal Processing
日期:2020-10-21
卷期号:: 1147-1152
被引量:2
标识
DOI:10.1109/wcsp49889.2020.9299765
摘要
Unmanned Aerial Vehicle (UAV) Edge Computing is expected to be critical for providing communications, computation, and storage services at areas with weak infrastructures through extending cloud service installed on low cost and easy-to-deploy UAVs to network edge. However, the service availability and capacity of one single UAV edge server is very limited, and can not meet the requirements of a number of mobile terminals (MTs) distributed in a large area. Therefore, it is more promising to have multiple UAVs collaborate with each other to provide edge computing service. In this circumstance, this paper establishes a Multi-UAV collaborative edge computing framework, in which the offloaded tasks to the UAVs from MTs are collaboratively processed through inter-UAV task offloading. Then, an optimization problem is built to minimize the system's energy consumption while completing all offloaded tasks. To solve this problem, we jointly optimize the number of deployed UAVs and the offloading decision at each MT. Firstly, a multi-UAV deployment mechanism based on differential evolution (DE) is adopted to determine each deployed UAV's position. Secondly, an efficient collaborative greedy algorithm for task scheduling is designed to help each MT decide whether to offload and its offloading destination. These two steps are iteratively conducted to reduce the number of deployed UAVs and system's energy consumption. To evaluate our proposal's performance, a series of simulations are conducted. Simulation results have shown that our proposal outperforms existing works in in terms of energy consumption as well as the number of deployed UAVs.
科研通智能强力驱动
Strongly Powered by AbleSci AI