摘要
针对偶发实时系统中过载处理资源消耗大的问题,提出一种基于最小可推迟时间和非精确计算的过载控制策略。结合最早截止期优先调度算法,利用最大挪用时间与最小可推迟时间动态判断系统负载状态,根据系统负载状态舍弃部分非重要任务,解决系统过载问题。实验结果表明,该策略可缩短过载处理时间,提高系统资源利用率。
Aiming at the problem that overload processing consumes high system resources in sporadic real-time systems,an overload control strategy based on Minimum Delay Time and Imprecise Computation(MDTIC) is proposed.Combined with the Earliest Deadline First(EDF) scheduling algorithm,the system load state is dynamically determined by using the maximum stealing time and minimum delay time,then some non-critical tasks are discarded according to the system load state to solve the system overload problem.Experimental results show that the proposed strategy can reduce the overload processing time and improve the utilization rate of the system resource.
作者
崔丽群
郭相卓
郭军
黄迪文
CUI Liqun;GUO Xiangzhuo;GUO Jun;HUANG Diwen(School of Software,Liaoning Technical University,Huludao,Liaoning 125100,China)
出处
《计算机工程》
CAS
CSCD
北大核心
2019年第6期108-114,共7页
Computer Engineering
基金
国家自然科学基金(61172144)
辽宁省教育厅科研项目(L2012113)