Research Article
Connectivity-Aware Virtual Machine Placement in 60 GHz Wireless Cloud Centers
@INPROCEEDINGS{10.1007/978-3-319-69605-8_4, author={Linghe Kong and Linsheng Ye and Bowen Wang and Xiaofeng Gao and Fan Wu and Guihai Chen and M. Hossain}, title={Connectivity-Aware Virtual Machine Placement in 60 GHz Wireless Cloud Centers}, proceedings={Cloud Computing, Security, Privacy in New Computing Environments. 7th International Conference, CloudComp 2016, and First International Conference, SPNCE 2016, Guangzhou, China, November 25--26, and December 15--16, 2016, Proceedings}, proceedings_a={CLOUDCOMP}, year={2017}, month={11}, keywords={}, doi={10.1007/978-3-319-69605-8_4} }- Linghe Kong
Linsheng Ye
Bowen Wang
Xiaofeng Gao
Fan Wu
Guihai Chen
M. Hossain
Year: 2017
Connectivity-Aware Virtual Machine Placement in 60 GHz Wireless Cloud Centers
CLOUDCOMP
Springer
DOI: 10.1007/978-3-319-69605-8_4
Abstract
Benefiting from the 60 GHz technology, physical machines in advanced cloud centers are connected by not only the conventional wired links but also the wireless communications. The 60 GHz (mmWave) introduces valuable advantages into cloud centers including flexibility, scalability and high rate. Nevertheless, mmWave is constrained by directional communications, i.e., a wireless link is connected if and only if two directional antennas face to each other. This constraint introduces a new problem in cloud service: the (VM) placement should consider the real-time connectivity if communications are required between VMs. Otherwise, rotating the antenna costs additional delay, resulting in performance degradation. To address this problem, we propose a novel (CAVMP) specialized for 60 GHz wireless cloud center. The core of CAVMP is to dynamically place VMs in order to improve the utilization and avoid overloads while taking the connectivity state into account. We build a 2-rack cloud to measure the connectivity feature of mmWave communications. In addition, we conduct extensive simulations to evaluate CAVMP. Performance results demonstrate that CAVMP significantly outperforms existing VM placement schemes in wireless cloud center.
