混合关键度驱动的实时调度研究

针对当前运行于不可预测开放环境下的嵌入式多使命复杂关键型系统,需要减少运行成本和处理不可预测工作负载情况的问题,文章提出一种混合关键度驱动的非对称式过载保护最小空闲调度策略。系统过载时,为共享同一处理器的不同关键度任务提供非对称式保护,禁止低关键度任务干扰高关键度任务,完全避免了传统的"关键度反转"问题。在恢复暂时阻塞的不同关键度任务时,在速率单调调度的基础上,引入关键度主导的截止期驱动动态调度策略,可使潜在处理器利用率达到100%。实验结果表明,这种新算法的综合性能优于当前已有的混合关键度任务调度算法。

An Effective Algorithm of Mixed-Criticality Driven Real-Time Scheduling

In many cases unexpected workload spikes are likely to occur due to unpredictable changes in the physical environment.In this paper we present a mixed-criticality driven asymmetric overload-protected shortest slack scheduling algorithm that implements an alternative protection scheme to avoid the criticality inversion problem.Sections 1 through 4 of the full paper explain our algorithm mentioned in the title,which we believe is new and effective and whose core consists of: "This algorithm can be used with rate monotonic based preemptive scheduler with deadline driven scheme to resume the blocked tasks.The potential processor utilization of the new algorithm can reach 100%.Section 1 briefs relevant past research.Section 2 deals with the model of the objects to be scheduled and relevant semantics.Section 3 deals with the problem of criticality inversion.Section 4 deals with our algorithm mentioned in the title." The experimental results,presented In Figs.2 through 5 and Table 2,show preliminarily that our new algorithm provides performance higher than those of two existing algorithms for mix-criticality task scheduling.