Using dynamic priority time Petri nets for scheduling analysis via earliest deadline first policy

In a previous paper [14], we investigated the expressiveness of Time Petri Nets extended with dynamic Priorities and showed that it is able to analyze the schedulability of a partitioned real-time system over a multiprocessor architecture. The scheduling analysis is proven via the Least Laxity First (LLF) policy and a set of independent tasks. In the same vein, we investigate in the present paper the applicability of the proposed approach for the Earliest Deadline First (EDF) policy to reduce the cost of preemption and we tackle the dependent tasks problem. Through an experiment, we present the capacity of our approach to detect the temporal faults in scheduling.