当前位置: X-MOL 学术Trans. Emerg. Telecommun. Technol. › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
An energy efficient and resource-constrained scheduling framework for smart city application
Transactions on Emerging Telecommunications Technologies ( IF 2.5 ) Pub Date : 2020-07-09 , DOI: 10.1002/ett.4040
Weipeng Jing 1 , Qiucheng Miao 1 , Houbing Song 2 , Yaqiu Liu 1
Affiliation  

With the development of smart Internet of things devices, intelligent applications are expected to lead further innovation in smart city. However, although cloud computing infrastructure can be used to meet traditional challenges, the scheduling model for new big data intelligent application has still not matured. In this work, we proposed a two-stage scheduling framework for smart city intelligent application. In the first stage, we propose a virtual machine selection algorithm for edge computing to enhance relative migration benefits. The algorithm defines the invalid virtual machine migration and relative migration benefits from the change in the overall computing resources of the cloud data center after the virtual machine migration. In the second stage, we proposed an energy efficient and resource-constrained scheduling framework for edge computing. The historical data of the cloud and edge computing workload of the computing node are processed in a sliding window manner, and the median absolute deviation of the historical data is used as the base of the physical reserved resource constraint when the base also changes as the workload changes. The experimental results show that energy-efficient and resource-constrained can make the computer resource provide high-quality services for users in a low-energy state.

中文翻译:

一种用于智慧城市应用的节能且资源受限的调度框架

随着智能物联网设备的发展,智能应用有望引领智慧城市的进一步创新。然而,虽然云计算基础设施可以应对传统挑战,但新的大数据智能应用的调度模型还没有成熟。在这项工作中,我们提出了智慧城市智能应用的两阶段调度框架。在第一阶段,我们提出了一种用于边缘计算的虚拟机选择算法,以增强相对迁移的好处。该算法定义了无效虚拟机迁移和虚拟机迁移后云数据中心整体计算资源变化带来的相对迁移收益。在第二阶段,我们为边缘计算提出了一种节能且资源受限的调度框架。计算节点的云和边缘计算工作负载的历史数据采用滑动窗口方式处理,当基数也随着工作负载变化时,以历史数据的绝对偏差中位数作为物理预留资源约束的基数变化。实验结果表明,节能和资源约束可以使计算机资源在低能耗状态下为用户提供高质量的服务。当基数也随着工作量的变化而变化时,以历史数据的中值绝对偏差作为物理预留资源约束的基数。实验结果表明,节能和资源约束可以使计算机资源在低能耗状态下为用户提供高质量的服务。当基数也随着工作量的变化而变化时,以历史数据的中值绝对偏差作为物理预留资源约束的基数。实验结果表明,节能和资源约束可以使计算机资源在低能耗状态下为用户提供高质量的服务。
更新日期:2020-07-09
down
wechat
bug