摘要
研究针对云计算平台以虚粒机为测量粒度,使用灰色关联度分析确定主要运行状态参数,以Hypervisor针对虚拟资源进行运行状态参数采样,使用长短时记忆(LSTM)循环神经网络建立虚拟机能耗模型用以度量云计算平台实时能耗。能耗度量采用WordCount与Sort进行节点能耗测量,得出能耗模型的实时平均误差为0.0677。
Based on cloud computing platform,the particle size is measured by virtual granulator.Main operating state parameters are determined by using grey relational analysis.Hypervisor is used for sampling of running state parameters of virtual resources.Long short time memory(LSTM)cyclic neural network is used to build virtual machine energy consumption model to measure real-time energy consumption of cloud computing platform.Energy consumption is measured by Word Count and Sort,and the real-time mean error of the energy consumption model is 0.0677.
作者
陈俊
张芥
田红珍
宋春香
邹文霞
CHEN Jun;ZHANG Jie;TIAN Hongzhen;SONG Chunxiang;ZOU Wenxia(School of Education,Guizhou Normal University,Guiyang 550025,China)
出处
《传感器与微系统》
CSCD
2020年第9期16-19,23,共5页
Transducer and Microsystem Technologies
基金
国家自然科学基金资助项目(61309006)
贵州师范大学资助博士科研项目([2013]3-21号)
贵州省教育厅高校人文社会科学研究项目(2020GH015)。
关键词
云计算平台
虚拟机
灰色关联度分析
长短时记忆(LSTM)
能耗模型
cloud computing platform
virtual machine
grey relational analysis
long short time memory(LSTM)
energy consumption model
