Reliability-oriented scheduling for static-priority real-time tasks in standby-sparing systems

Abstract

The advent of complicated embedded systems with regard to relentless technology scaling and integration of more components into a single chip, have caused these systems to be less reliable. Moreover, these advancements have accompanied with a drastic increase in energy consumption. However, using energy management techniques such as Dynamic Voltage and Frequency Scaling (DVFS) has made reliability issues to be exacerbated. Hence, energy efficiency and fault-tolerance are two conflicting key objectives in the design of efficient real-time embedded systems. In this paper by considering periodic real-time tasks in standby-sparing system, we propose a novel online scheduling technique, which tackles this problem. The aim of this paper is twofold: (1) to show that scheduling primary and backup tasks in different processors in a moderate speed for static-priority real-time tasks will lead to better energy savings; (2) to explore the effectiveness of this technique for mixed-criticality tasks. Our simulation results reveal a significant energy saving (up to 25%) in comparison with the state-of-the-art scheme based on standby-sparing system, while preserving system's original reliability.