异构网格环境下的可信依赖任务调度

异构网格环境下的可信依赖任务调度 Trusted Scheduling of Dependent Tasks under Heterogeneous Grid Environment 朱海 ZHU Hai 1 first-author 王宇平 WANG Yuping 1 王晓丽 WANG Xiaoli 1 范磊 FAN Lei 1 西安电子科技大学计算机学院 西安电子科技大学计算机学院 School of Computer Science and Technology,Xidian University School of Computer Science and Technology,Xidian University 针对异构网格环境下依赖任务调度过程中网格节点行为可信性考虑不足的问题,根据网格节点的历史行为表现,构建了一个动态信誉度评估策略,为确立任务需求与资源节点行为可信属性之间的隶属关系,定义了隶属度函数,建立了一种安全可信的网格任务调度新模型.为了实现该模型,提出一种依赖任务可信调度的粒子群进化算法.该算法通过深度值和关联耦合度的引入解决了任务间的约束关系;为克服传统粒子群算法解决离散问题时的不足,结合网格任务调度问题的具体特点,重新定义并设计了新的粒子进化方程;为预防算法陷入局部最优,引入了均匀扰动速度.仿真实验表明,该算法与同类算法相比,在相同条件下具有较高的执行效率和安全可信优化性能等. In order to solve the dependent tasks scheduling problem which did not take grid nodes behavior credibility into consideration under the heterogeneous grid environment,dynamic credibility assessment strategies are constructed according to the history of grid nodes behavior.Meanwhile,in order to establish the subordinate relationship between the tasks requirements and security attributes of the resource nodes behaviors,membership functions are defined.A new model of security reliable grid task scheduling is proposed.In order to solve this model,a credible task scheduling particle swarm evolutionary algorithm is proposed.In this algorithm,first,through the introduction of the depth value and the associated coupling degree,the constraint relationship among tasks is solved.Second,in order to overcome the inadequency of the traditional particle swarm algorithm to solve the discrete problems,a new particle evolution equation is re-defined and re-designed by considering the specific characteristics of grid task scheduling problem.Finally,a uniform speed disturbance is proposed in order to prevent the algorithm into local optimum.Simulation results show that this algorithm has a higher efficiency and a better performance of trusted security optimization in comparison with similar algorithms under the same conditions. 网格计算 依赖任务调度 可信优化模型 粒子群进化算法 grid computing dependent tasks scheduling trusted optimization model particle swarm evolution algorithm 10.14188/j.1671-8836.2010.02.025 TP393.01 国家自然科学基金资助项目（60873099） 2010-02-01 2021-04-01 2 56