多目标约束的网格任务安全调度模型及算法研究

朱海; 王宇平

doi:10.3724/SP.J.1146.2009.00634

多目标约束的网格任务安全调度模型及算法研究

doi: 10.3724/SP.J.1146.2009.00634

朱海,
王宇平

计量
- 文章访问数: 3882
- HTML全文浏览量: 121
- PDF下载量: 774
- 被引次数: 0
出版历程
- 收稿日期: 2009-04-28
- 修回日期: 2009-09-30
- 刊出日期: 2010-04-19

Constrained Multi-objective Grid Task Security Scheduling Model and Algorithm

摘要

摘要: 异构网格环境的特点决定了其任务调度是受调度长度、安全性能及调度费用等多个因素制约的。该文根据网格资源调度的特点构造了一个安全效益函数和节点信誉度动态评估模型，并以此为基础建立了一个多目标约束的网格任务调度模型。利用隶属度函数将多目标函数转化为单目标模型，通过设计新的进化算子，从而提出一种遗传算法MUGA(Mode Crossover and Even Mutation Genetic Algorithm)进行求解，并对算法的收敛性进行了理论分析。仿真实验表明，在同等条件下该算法与同类算法相比，在任务调度长度、安全效益值、可信度及调度费用指标优化方面具有较好的综合性能。
- 异构网络; 网格计算; 任务调度; 安全效益函数; 信誉度; 遗传算法
Abstract: The characteristic of heterogeneous grid environment determines that the task scheduling is constrained by a number of factors such as the length of scheduling, the performance of security, the cost of scheduling and etc. Firstly, based on the characteristics of grid task scheduling, a security benefit function and an efficient nodes credibility dynamic evaluation model are constructed. Then a constrained multi-objective grid task scheduling model is proposed. Secondly, by using the subjection degree function, the multi-objective optimization is transformed into a single objective optimization issue. Thirdly, Through the design of new evolutionary operators, a new genetic algorithm is proposed. The convergence of this algorithm is analyzed. Simulation results show that the proposed algorithm is better than the compared ones in terms of the length of the task scheduling, security efficiency value, reliability and scheduling costs.

HTML全文

参考文献(1)

Chakrabarti A, Damodaran A, and Sengupta S. Gridcomputing security: A taxonomy[J].IEEE Security Privacy.2008, 6(1):44-51[2]Song S S, Hwang Kai, and Kwok Yu-kwong. Risk-resilientheuristics and genetic algorithms for security-assured grid jobscheduling [J].IEEE Transactions on Computers.2006, 55(6):703-719[3]Braun T D, Slegel H J, and Becj N. A comparison of elevenstatic heuristics for mapping a class of independent tasks ontoheterogeneous distributed computing systems [J].Journal ofParallel and Distributed Computing.2001, 61(6):810-837[4]Ranaldo N and Zimeo E. Time and cost-driven scheduling ofdata parallel tasks in grid workflows[J].IEEE SystemsJournal.2009, 3(1):104-120[5]He X, Sun X, and Laszewski G V. QoS guided min-minheuristic for grid task scheduling [J].Journal of ComputerScience and Technology.2003, 18(4):442-451[6]Yi K and Wang R C. Nash equilibrium based task schedulingalgorithm of multi-schedulers in grid computing[J]. ActaElectronica Sinica, 2009, 37(2): 329-333.[7]Azzedin F and Maheswaran M. Integrating trust into gridresource management systems[C]. Proceedings of 2002International Conference on Parallel Processing. LosAlamitos, Cal, USA: IEEE Computer Society Press, 2002:47-54.[8]Zhu H and Wang Y P. Security-Driven Task SchedulingBased on Evolutionary Algorithm[C]. Proceeding of theComputational Intelligence and Security 2008, IEEE Press,2008: 451-456.[9]张伟哲, 胡铭曾, 张宏莉等. 多QoS 约束网格作业调度问题的多目标演化算法[J]. 计算机研究与发展, 2006, 43(11):1855-1862.Zhang W Z, Hu M Z, and Zhang H L, et al.. A multi-objectiveevolutionary algorithm for grid Job scheduling of Multi-QoSconstraints[J]. Journal of Computer Research andDevelopment, 2006, 43(11): 1855-1862.[10]Li B B and Wang L. A hybrid quantum-inspired geneticalgorithm for multi-objective flow shop scheduling [J].IEEETransactions on Systems, Man, and Cybernetics.2007, 37(3):576-591[11]张伟哲, 方滨兴, 胡铭曾等. 基于信任QoS 增强的网格服务调度算法[J]. 计算机学报, 2006, 29(7): 1157-1166.Zhang W Z, Fang B X, and Hu M Z, et al.. A trust-QoSenhanced grid service scheduling[J]. Chinese Journal ofComputers, 2006, 29(7): 1157-1166.

施引文献

资源附件(0)

访问统计