Execution time budget assignment for mixed criticality systems

Indeed, the methods that use the are executed on the same processor. The challenge is full distribution to compute the probabilistic response time that low criticality tasks do not disturb the good functioning have a high complexity for exact methods or have to make of the high criticality ones. In real-time scheduling, since the assumptions on the shape of the distributions for analytical original Vestal's model [1], a classical model has emerged, see methods. The contributions of the paper are: (1) We propose [2] for a complete survey. In this model, tasks have several a definition of execution time variability and a method for its execution times budgets, one budget per possible criticality. If quantification using statistical dispersion parameters, (2) We a task does not signal its termination after the execution of propose a heuristic that uses the execution time variability to its allocated budget at a certain criticality level, the system solve the scheduling problem of a mixed criticality system, (3) moves to the next criticality level. In every system criticality We evaluate our approach using simulations and benchmarks level, only tasks of criticality equal or higher to the criticality executed on an ARM-Cortex A53. of the system have to respect their deadlines.