Mixed-Criticality Scheduling of Energy-Harvesting Systems

Abstract

Energy harvesting is a promising approach to powering real-time embedded devices which are deployed wherever it is not possible or practical to recharge. Since the stochastic nature of harvested energy makes it challenging to simultaneously guarantee both timing and energy constraints of energy-harvesting real-time systems, the worst-case performance analysis becomes more crucial when analyzing the system schedulability. In this paper, we study the performance analysis problem of energy-harvesting mixed-criticality (EHMC) systems scheduled by an energy-aware adaptation of EDF. In particular, we propose a new method that can be used to derive time demand bounds for a mixed-criticality task set, which upper-bound the total amount of time required to satisfy both the processor and energy demand of the task set in any time interval of a given size for each criticality mode. Moreover, we calculate the minimum size of the capacitor for our schedulability test to be valid. Experiment results show that our approach is significantly more powerful than previous approaches to energy harvesting mixed-criticality systems.