期刊
IEEE TRANSACTIONS ON SUSTAINABLE COMPUTING
卷 7, 期 2, 页码 313-323出版社
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/TSUSC.2018.2878438
关键词
Servers; Power demand; Cloud computing; Mobile handsets; Task analysis; Computer architecture; Load management; 5G networks; mobile edge computing; hotspot mitigation; energy saving
类别
资金
- Ministry of Science and Technology of Taiwan, ROC [MOST 106-2221-E-005-015, MOST 107-2221-E-005-017]
Mobile edge computing (MEC) provides cloud computing capabilities in close proximity to mobile users, leading to better performance and reduced power consumption. However, the varying workloads of MEC servers result in load imbalance and hotspot issues. To address this, we propose a scheme called COHM, which considers the computation tasks for MEC servers with different states to mitigate hotspots.
Due to the long latency of accessing services from cloud datacenters, mobile edge computing (MEC) has been proposed to provide the capability of cloud computing in close proximity to mobile users. Mobile devices can offload computation-intensive applications to MEC servers to achieve better performance and reduce power consumption. Owing to user mobility, workloads of MEC servers would vary severely to cause imbalanced loads and incur hotspot servers. Previous studies of power saving and hotspot management apply mechanisms for cloud datacenters to MEC servers, but these studies may not be suitable for MEC servers due to the limited computation and communication capacity. We are motivated to propose a scheme, complementary offloading for hotspot mitigation (COHM), for MEC. COHM consists of mechanisms for MEC servers with different states, namely busy, idle, and available. These mechanisms jointly consider the computation tasks for MEC servers with different states to mitigate hotspots. The simulation results show that COHM can decrease the number of hotspot servers without degrading the quality of service for applications.
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