Probabilistic analysis for mixed criticality systems using fixed priority preemptive scheduling

International audienceThis paper introduces probabilistic analysis for fixed priority preemptive scheduling of mixed criticality systems on a uniprocessor using the Adaptive Mixed Criticality (AMC) and Static Mixed Criticality (SMC) schemes. We compare this analysis to existing deterministic methods, highlighting the performance gains that can be obtained by utilising more detailed information about worst-case execution time estimates described in terms of probability distributions. Besides improvements in schedulability, we also demonstrate signiicant gains in terms of the budgets that can be allocated to LO-criticality tasks. A preliminary version [26] of the research described in this paper was published in the Workshop on Mixed Criticality Systems (WMC) in 2016. In this paper, we correct the analysis given in [26], ensuring that the schedulability of HI-criticality tasks does not depend on the behavior of LO-criticality tasks. Further, we provide an alternative analysis (in Section 4.4) and show how support for LO-criticality tasks can be improved via increased execution time budgets (in Section 4.6)