通信竞争的混合关键级系统多DAG动态调度策略

发布时间：2018-05-28 21:52

本文选题：通信竞争 + 混合关键级系统　；参考：《计算机研究与发展》2015年11期

【摘要】：以多DAG模型研究通信竞争的混合关键级系统(mixed-criticality systems)的调度问题是适应现代汽车电子系统异构化和分布式的需要.首先实现通信竞争环境下"向上排序值(upward rank value)"和"最早完成时间(earliest finish time)"中时间的精确分析,以适应系统中计算与网络均异构,且任务与消息的同步特征.接着提出公平策略的多DAG动态任务与消息调度F_MDDTMS算法,以降低系统的调度长度;提出关键级策略的多DAG动态任务与消息调度C_MDDTMS算法,以确保高关键级应用的实时性;结合F_MDDTMS算法和C_MDDTMS算法,提出混合关键级策略的多DAG动态任务与消息调度MC_MDDTMS算法,既确保混合关键级系统中高关键级应用的实时性,又使得低关键级应用得到积极的处理.实例分析和实验结果验证了提出的算法在调度长度、不公平性、最差响应时间和实时性上的优越性.
[Abstract]:The scheduling problem of mixed-criticality systems for communication competition based on multi-DAG model is suitable for the isomerization and distribution of modern automotive electronic systems. First, the accurate analysis of the time in "upward rank value" and "earliest complete finish time" in communication competition environment is realized to adapt to the isomerism of computing and network in the system, and the synchronous characteristics of tasks and messages. Then, a fair policy multi-DAG dynamic task and message scheduling F_MDDTMS algorithm is proposed to reduce the scheduling length of the system, and a key-level policy multi-DAG dynamic task and message scheduling C_MDDTMS algorithm is proposed to ensure the real-time performance of high critical level applications. Combined with F_MDDTMS algorithm and C_MDDTMS algorithm, this paper proposes a hybrid key-level multi-level dynamic task and message scheduling MC_MDDTMS algorithm, which not only ensures the real-time performance of high critical level applications in hybrid key-level systems, but also makes the low-critical level applications actively processed. The example analysis and experimental results show the superiority of the proposed algorithm in scheduling length, unfairness, worst response time and real-time performance.
【作者单位】：湖南大学信息科学与工程学院;湖南大学嵌入式与网络计算湖南省重点实验室;湖南省发展和改革委员会;厦门理工学院计算机与信息工程学院;
【基金】：国家自然科学基金项目(61173036,61202102,61300039,61300037,61502405) 国家“八六三”高技术研究发展计划基金项目(2012AA01A301-01)
【分类号】：U463.6;TP368.1

【参考文献】