Energy and quality of control constraints in real-time scheduling of synchronous hybrid tasks

Abstract

The embedded systems are characterized by their autonomic functioning whose energy supply is ensured by batteries. Therefore, the reduction of their power consumption and more specifically, the quality of their control becomes the crucial metric optimization in the design of such systems. In this paper, we investigate the integration of precedence, resources sharing and quality of control constraints in the real time scheduling of firm periodic and aperiodic tasks. To study this problem, we start by adapting the analytical conditions of scheduling having been proposed in literature. We also treat the problem of dynamic voltage scaling of the processor in the same context. The community of the researchers have proposed many dynamic voltage-scaling algorithms. However, these algorithms do not consider the precedence, shared resources and quality of control constraints on the scheduling of firm periodic and aperiodic tasks. In this case, we propose two algorithms with the constraints cited above. These algorithms are based on the analytical model adopted in this paper. Experimental results show that the proposed algorithms reduce the energy consumption under earliest deadline first scheduling policy and the stack resource protocol.