Optimizing Energy in Non-Preemptive Mixed-Criticality Scheduling by Exploiting Probabilistic Information

Abstract

The strict requirements on the timing correctness biased the modeling and analysis of real-time systems towards the worst-case performances. Such focus on the worst-case, however, does not provide enough information to effectively steer the resource/energy optimization. In this paper, we integrate a probabilistic-based energy prediction strategy with the precise scheduling of mixed-criticality tasks, where the timing correctness must be met for all tasks at all scenarios. The Dynamic Voltage and Frequency Scaling (DVFS) is applied to this precise scheduling policy to enable energy minimization. We propose a probabilistic technique to derive an energy-efficient speed (for the processor) that minimizes the average energy consumption, while guaranteeing the (worst-case) timing correctness for all tasks, including lo-criticality ones, under any execution condition. We present a response time analysis for such systems under the nonpreemptive fixed-priority scheduling policy. Finally, we conduct an extensive simulation campaign based on randomly generated task sets to verify the effectiveness of our algorithm (w.r.t. energy savings) and it reports up to 46% energy-saving.

MLCAD: A Survey of Research in Machine Learning for CAD Keynote Paper

“…The group also used hierarchical MDPs [141] to efficiently model both the communication environment and the computing system platform. Bhuiyan et al [142] described a probabilistic approach to energy-optimized scheduling of multicriticality systems. Their multicriticality model includes two criticality levels: 1) HI and 2) LO.…”

A Mixed-Criticality Approach to Fault Tolerance: Integrating Schedulability and Failure Requirements

“…To deal with mode switches, traditional methods drop tasks of lower criticality than the WCET criticality of the task that overran. However, several discussions emerged recently on this strategy [7]- [9], which is considered far from the industry perspective. In fact, dropping tasks during mode-switch creates a "dependency" between the tasks, i.e., the functional correctness of a job of the task does not depend on the job itself, but it is influenced by the execution of other tasks.…”

Beyond the Traditional Analyses and Resource Management in Real-Time Systems

“…In this way, an incorrect estimation of the distribution does not invalidate the safety properties. This is an active area of research and subject of a recent work [2].…”