摘要
为解决混合云环境下科学工作流数据布局问题,在考虑数据的安全需求的前提下,以优化跨数据中心传输时延为目标,提出了一种混合云环境下面向安全的科学工作流布局策略。分析数据集的安全需求以及数据中心所能提供的安全服务,提出安全等级分级规则;设计并提出基于遗传算法和模拟退火算法的自适应粒子群优化算法(adaptive particle swarm optimization algorithm based on SA and GA,SAGA-PSO),避免算法陷入局部极值,有效提高种群多样性;与其它经典布局算法对比,基于SAGA-PSO的数据布局策略在满足数据安全需求的同时能够大大降低传输时延。
To address the issue of data layout in scientific workflows in a hybrid cloud environment,while considering the security requirements of the data,a security-oriented optimization strategy was proposed for scientific workflow layout with the goal of optimizing the cross-data center transmission delay.The security requirements of the dataset and the security services that could be provided by the data center were analyzed,and a security classification rule was proposed.An adaptive particle swarm optimization algorithm based on genetic algorithm and simulated annealing algorithm(SAGA-PSO)was designed and proposed to avoid the algorithm from getting stuck in local optima and effectively improve population diversity.Compared with other classic layout algorithms,the data layout strategy based on SAGA-PSO can significantly reduce transmission delay while meeting the data security requirements.
作者
苏明辉
林兵
卢宇
王素云
SU Ming-hui;LIN Bing;LU Yu;WANG Su-yun(College of Physics and Energy,Fujian Normal University,Fuzhou 350117,China;School of Electronics Engineering and Computer Science,Peking University,Beijing 100871,China;Concord University College,Fujian Normal University,Fuzhou 350117,China)
出处
《计算机工程与设计》
北大核心
2024年第7期2004-2012,共9页
Computer Engineering and Design
基金
国家重点研发计划基金项目(2018YFB1004800)
国家自然科学基金项目(62072108、61672159)
福建省高校产学合作基金项目(2022H6024)
福建省自然科学基金项目(2019J01244)。
关键词
混合云
科学工作流
数据布局
安全分级
时延优化
遗传粒子群优化算法
模拟退火
hybrid cloud
scientific workflow
data placement
security classification
delay optimization
genetic particle swarm optimization algorithm
simulated annealing
