云环境下多核仿真平台虚拟任务数据索引  

Virtual Task Data Index of Multi Core Simulation Experiment Platform in Cloud Environment

作　　者：陶建平 曹霞 TAO Jian-ping;CAO Xia(Network and Computing Center,Huazhong University of Science and Technology,Wuhan Hubei 430074,China)

机构地区：[1]华中科技大学网络与计算中心,湖北武汉430074

出　　处：《计算机仿真》2021年第11期389-393,共5页Computer Simulation

基　　金：华中科技大学实验技术研究项目(2021);华中科技大学2020年教学研究项(2020172)。

摘　　要：以优化多核仿真平台虚拟任务数据管理效率为目标,提出云环境下多核仿真平台虚拟任务数据索引设计方法。在云环境下,基于分形维数的选择性聚类融合,有效聚类各类型数据;依据虚拟任务数据索引机制,将虚拟任务的各类型数据变换为一维索引,将一维索引前缀设成平台主键数据,体现数据在覆盖网络结构中的所在位置,实现虚拟任务数据索引的设计。研究结果显示:当虚拟任务数据中心规模、查询请求维数、查询区间出现变化时,上述方法应用后,数据管理效率最高;且在不同并发用户人数条件下该方法的应用效果也优于其它方法。A method of virtual task data index based on cloud environment was designed for optimizing the efficiency of virtual task data management of multi-core simulation experiment platform.In the cloud environment,all kinds of data were clustered by the selective clustering fusion based on fractal dimension.The data index mechanism of virtual task was adopted to convert all kinds of data of virtual task into one-dimensional index.Then,the prefix of the one-dimensional index was set as the primary key data of the platform.The location of data in overlay network structure was presented,and the design of virtual task data index was completed.The results show that this method has high data management efficiency,and the application effect of this method is better than that of traditional methods under the condition of different number of concurrent users.

关 键 词：云环境 多核 仿真平台 虚拟任务 数据索引 聚类融合 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]