摘要
为了解决LSF调度算法在实时调度中由颠簸现象引起的调度实时性差、浪费系统资源的问题,在LSF算法中引入一个任务重要度系数,采用云模型对任务重要度系数和裕度进行定量表示,并通过由重要度系数云和裕度云两个任务特征参数云模型共同确定的二维云模型,为每个任务设定一个抢占阈值,当某一就绪任务要抢占当前任务时,必须要满足它的优先级高于当前任务的抢占阈值。仿真结果表明,采用云模型优化后的LSF算法不仅有效解决了颠簸现象,而且能使紧急且重要的任务优先运行。
In order to solve the problem of poor real-time and system resources waste,which caused by jounce phenomenon of the LSF scheduling algorithm in real-time scheduling,the weighty coefficient of the task is introduced in the LSF algorithm.Using cloud model,the weighty coefficient and the slack time are quantificationally represented.Each task is assigned a preemption threshold through a twodimensional cloud model which codetermined by the importance coefficient cloud and the slack time cloud.When a task is ready to seize the current task,its priority must be higher than the preemption threshold of the current task.Simulation results show that the optimized LSF algorithm not only effectively solve the jounce phenomenon but also make the most urgent and important task run first.
出处
《计算机工程与设计》
CSCD
北大核心
2010年第13期3014-3016,3042,共4页
Computer Engineering and Design
基金
国家自然科学基金项目(60964003)
甘肃省科技攻关基金项目(2GS052-A52-003-11)
关键词
云模型
重要度
裕度
抢占阈值
实时性
cloud model
importance factor
slack time
preemption threshold
real-time
作者简介
作者简介：陈旭辉（1972-），男，陕西户县人，博士，副教授，研究方向为人机交互、嵌入式系统、新一代互连网技术；
于国龙（1981-），男，辽宁东港人，硕士研究生，研究方向为嵌入式系统。E-mail：xhchen@lut．cn
