期刊
IEEE INTERNET OF THINGS JOURNAL
卷 6, 期 2, 页码 3455-3466出版社
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/JIOT.2018.2885453
关键词
Cloud networks; computation resources; Internet of Things (IoT); mobile edge server; profit maximization
资金
- Natural Sciences and Engineering Research Council of Canada
- National Natural Science Foundation of China [61702450, 61629302]
- CSC (China Scholarship Council)
- Compute Canada
We study the computation resource management problem in mobile edge-cloud computing networks. Mobile edge servers shall first satisfy the computation requirements of mobile users and Internet of Things (IoT) devices, and then wholesale redundant computation resources to the cloud networks to maximize their profit. Due to the coarse time granularity of wholesales, computation resource buyback may happen occasionally to deal with traffic bursts. Thus, the mobile edge servers need to make a tradeoff between the wholesale profit and the buyback cost. In this paper, the computation resource management problem is modeled as profit maximization. To solve this problem, we first analyze the relationship among the reserved computation resources, the computation tasks of mobile users and IoT devices, and the buyback cost. Then, we design an efficient wholesale scheme to determine the amount of the wholesaled computation resources, by which the total expected profit of the mobile edge server can be maximized. Given the reserved computation resources, we also propose a fast-convergent realtime buyback scheme for mobile edge servers to minimize the buyback cost. Finally, the simulation results show that our proposed efficient wholesale and buyback scheme can increase the total profit while guaranteeing the computation delay of all the computation tasks, especially when the computation workloads are time-varying.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据