面向安全关键实时应用的分层防危调度算法研究

Hierarchical Safeguard Scheduling Algorithm for Safety Critical Real-time Application

针对现有防危调度算法在软硬件失效情况下防危能力不足的问题,具体进行了以下工作:构建了一种分层防危实时调度模型,该模型从功能组件和安全分区两方面描述了安全关键实时应用的防危性需求,并给出一种基于分层调度思想的三级防危调度器框架。以该模型和框架为基础,提出了一种新的分层防危调度算法(HSS),该算法对安全关键实时应用中不同关键度的功能组件采用空间隔离机制,对同一功能组件内的不同分区采用时间隔离机制,兼顾实现了时空隔离的防危效果。仿真实验结果表明,HSS算法与其他同类算法相比,在防危效果和应用负载承受能力方面具有较好的表现。

The paper solved the problem that current popular safeguard scheduling algorithm cannot achieve the safeguard function under the environment of software and hardware failure.It built a new hierarchical real-time scheduling model which describes the safety requirement of safety-critical real-time application from two aspects including function component and safe partition,and designed a three-level safeguard scheduler framework.Based on the model and framework,the paper proposed a new hierarchical safeguard scheduling algorithm（HSS） which achieves spacial separation effect by distributing function components with different critical degrees to different physical processor clusters,and attains temporal separation effect by activating various partitions running on the same processor in a fixed cycle.Empirical investigations show that the improvements in the safeguard performance and the endurance to different application loads can be achieved by choosing HSS than other similar algorithms.