Hierarchical Modeling with dynamic Priority Time Petri Nets for Multiprocessor Scheduling Analysis

Dynamic Priority Time Petri Nets (dPTPN) represent a powerful formalism for the scheduling analysis of Real-Time Systems running on Multiprocessor architecture. The originality of the dPTPN semantics, compared to the existing research work, is the dynamic calculation of the priority of transitions in conflict. The present paper presents a new modeling strategy with dPTPN based on object modeling concept. Thus, a new component is proposed and the scheduling model is constituted with different instances of it. The scheduling is assured through the Earliest Deadline First with a set of dependent tasks. We prove the capacity of our approach to detect the non-schedulable sequences via an experiment.