Title: Size analysis in multiprocessor real-time scheduling
Author: Annie Choquet-Geniet; Gaëlle Largeteau-Skapin
LIAS, Université de Poitiers and ENSMA, BP 40109, 86961 Futuroscope Chasseneuil Cedex, France
Xlim-SIC, Université de Poitiers, BP 30179, 86962 Futuroscope Chasseneuil Cedex, France
Journal: Int. J. of Critical Computer-Based Systems, 2014 Vol.5, No.3/4, pp.197 - 217
Abstract: If timeliness is a strong requirement for real-time applications, the reduction of weight and of the energy consumption are also important concerns. Besides, the actual trend is towards multiprocessor platforms. Thus we are interested in the determination of the lowest number of processors for a real-time system. For that aim, we use a model-driven approach, based on a geometrical modelling, and the matrix calculus in the dioid (N; max; min). Then, we discuss the size of the schedule that must be computed for a full temporal validation of the system.
Keywords: schedule size; multiprocessor scheduling; real-time scheduling; geometrical modelling; dioid; resource optimisation; cyclicity; weight reduction; energy consumption; real-time applications.
Available online 03 Sep 2014