Sustainable Quality of Service for Real-Time Jobs in Autonomous Computing Devices

International audience; Energy harvesting has emerged as an efficient way for powering wireless devices in order to extend their lifetime. In this paper a dynamic power manage- ment strategy is described that guarantees the correct execution of every feasible application for a given energy harvester, energy storage unit and processor. The so-called ED-H strategy was recently proved to be optimal in that sense that ED-H optimizes the Quality of Service defined as the ratio of deadline success for the real-time jobs. We propose to describe this novel strategy for dynamic power management and scheduling.