Improving schedulability and energy efficiency for window-constrained real-time systems with reliability requirement

Abstract

For real-time embedded systems, schedulability, energy efficiency, Quality of Service (QoS), and reliability are four highly co-related important design concerns. In this paper, we explore combining these four dimensions of design issues to achieve better schedulability and energy efficiency for real-time systems while satisfying the QoS and reliability requirements. The QoS requirements are deterministically quantified with the window-constraints, which require that at least m out of each non-overlapped window of k consecutive jobs of a real-time task meet their deadlines. Our contributions mainly consists of two parts: in the first part, we propose two off-line approaches for the management of appropriate mandatory/optional job partitioning with the purpose of improving the schedulability for real-time systems with window-constraints; based on it, in the second part, we proposed advanced techniques to reduce the energy consumption of the real-time systems while satisfying the reliability requirements. The results of extensive experiments demonstrate that our proposed techniques significantly outperform previous approaches in both schedulability and energy efficiency for real-time embedded systems while satisfying the window-constraints and reliability requirements.