某演练评估系统中云计算任务调度研究  

Task Scheduling of Cloud Computing in Training Evaluation Simulation System

作　　者：张磊[1] 晁爱农[1] 郭利锋[1] 

机构地区：[1]空军工程大学工程学院,陕西西安710038

出　　处：《计算机仿真》2012年第7期114-116,134,共4页Computer Simulation

摘　　要：研究合同战术演练评估系统应用中的云计算任务调度问题。针对目前的云计算调度算法研究大都是基于通用性或者商业需求,对军事应用特点考虑不多,应用到合同战术演练评估系统中无法满足系统对于调度实时性等性能的要求的问题,通过分析云计算的任务调度特点,引入数据存储节点优先和节点效能的概念提出了一种改进的基于负载均衡的任务调度算法,算法减少了数据存取时间并采用节点效能的概念能更准确地描述主机性能。仿真结果验证了改进后的算法在任务数量增大时任务执行的速度有所提升,能更好地满足合同战术演练评估系统复杂度和规模增大对实时性的需求。Task scheduling of cloud computing in a combined-tactical training evaluation system was studied in this paper. Most of the current scheduling algorithm is based on universal or business needs and takes limited military needs into consideration. They can not meet the real-time requirements of tactical training evaluation system. By ana- lyzing the characteristics of cloud computing task scheduling, an improved task scheduling algorithm based on Load Balancing was proposed by introducing storage nodes priority and node Efficacy. Data access time was reduced which can more accurately describe the performance of the host in this algorithm. The result of simulation shows that it can improve the schedule speed when the number of tasks increases and can meet the demand better for the real-time re- quirements of the tactical training evaluation system for complexity and Large-scale.

关 键 词：合同战术演练 仿真 云计算 虚拟化 任务调度 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]