FMCI: An EDF-VD-based flexible mixed-criticality task scheduling method

Abstract

At present, mixed-criticality systems have become the development trend of embedded systems, such as in power edge computing, avionics safety control and car safety driving scenarios. In the existing advanced flexible mixed-criticality (FMC) model, the method of calculating compensation requirements may have a large number of low-criticality invalid compensation, which will greatly reduce the overall performance of low-criticality tasks. In response to this problem, our paper improves the method of reducing the service level of low-criticality tasks, from reducing the execution time of low-criticality tasks to now extending the period of low-criticality tasks, and proposes the corresponding FMCI-EDF-VD scheduling algorithm. Experimental results show that flexible mixed criticality improved (FMCI) model can manage the decline in service levels of low-criticality tasks more wisely and reduce invalid compensation for low-criticality tasks. Compared with existing FMC, our method can always perform better in ensuring the completion of high-criticality tasks while scheduling more low-criticality tasks under different circumstances.