In: Proceedings of 9th International Workshop on Petri Nets and Performance Models, PNPM'01 Aachen, Sept. 11-14, 2001, Reinhard German and Boudewijn Haverkort (eds.), IEEE, pages 125-134. 2001.
Abstract: Concurrent Generalized Petri Nets (CGPNs) are a subclass of Markov Regenerative Stochastic Petri Nets characterized by timed transitions, with general distributed firing times (GEN transitions), that become enabled simultaneoulsy while no other GEN transitions can be activated during their enabling periods. In their original definition, CGPNS are identified through the study of their state space, hence they are recognized only after the construction of their tangible reachability graphs. In this paper we present sufficient conditions representing a first step in the definition of an alternative method for the characterization of this type of models. The method is based on the structural analysis of the net, it does not require the generation of its tangible reachability graph, and is thus computationallly convenient. Moreover, an advantage of this method is that it can be applied also in case of models characterized by large state spaces, hence prone to be analyzed via simulations using regenerative techniques. The criteria that we propose represent the basis for a preliminary analysis of the model in order to verify its membership o the class of CGPNs whose numerical solution always requires the construction of its tangible reachability graph and the characterization of the underlying Markov regenerative process.