车辆移动云可靠性任务调度  

TASK SCHEDULING OF VEHICLE MOBILE CLOUD RELIABILITY

作　　者：李晓静[1] 马海英[2] Li Xiaojing;Ma Haiying(Jiyuan Vocational and Technical College,Jiyuan 459000,Henan,China;School of Computer Science and Technology,Nantong University,Nantong 226019,Jiangsu,China)

机构地区：[1]济源职业技术学院,河南济源459000 [2]南通大学计算机科学与技术学院,江苏南通226019

出　　处：《计算机应用与软件》2019年第11期78-85,共8页Computer Applications and Software

基　　金：国家自然科学基金项目(61402244,61371111);济源市科技攻关计划项目(16022017)

摘　　要：车辆移动性使得移动车辆云中的任务调度可靠性问题变得尤为复杂。针对这一问题,提出一种基于混合整数线性规化最优化方法的云任务调度算法。借助于MapReduce构建车辆云任务的调度模型,并设计一种复杂度更低的启发式调度方法,在有效降低任务执行延时的同时,确保了任务调度的可靠性。通过在网络仿真器NS3中运行城市道路环境下的MapReduce应用,对算法的调度结果进行性能评估。结果表明,与同类的车辆云中的调度方法相比,该算法在作业平均执行时间、作业调度成功率、系统吞吐量及任务执行开销等性能指标上均优于对比算法。The mobility of the vehicles makes the reliability of task scheduling in mobile vehicle cloud more complex.Aiming at this problem,this paper proposes a cloud task scheduling algorithm based on mixed integer linear programming optimization.To solve the problem of the resource heterogeneity and the computation parallelization in vehicle cloud task scheduling,the algorithm constructed a scheduling model of vehicle cloud task by means of MapReduce.And we designed a heuristic scheduling method with lower complexity to effectively reduce the tasks execution delay and guarantee the reliability of task scheduling.By applying MapReduce in urban road environment running in network simulator NS3,the performance of the algorithm was evaluated.The results show that,compared with the state-of-the-art scheduling methods in vehicle cloud,our algorithm is superior to the comparative algorithm in performance indicators such as average job execution time,job scheduling success rate,the system throughput and task execution overhead.

关 键 词：车辆云计算 任务调度 可靠性 启发式算法 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]