Subnetwork time distributions as a means for multilevel evaluation of generalized stochastic Petri nets

Abstract

A new hierarchical evaluation procedure for generalized stochastic Petri nets (GSPN) is being presented. It is based on extensions of flow equivalent aggregation (FEA). At every level of hierarchy, subnets are approximated by substitute networks. As criterion for the similarity of the networks, the subnetwork time distribution (STD) is used which is the sojourn time distribution of a token X in a subnet conditioned on the token distribution at the epoch of arrival of X and the context into which the net is embedded for determining the STD. The computation of a substitute network's type and parameters is outlined. The performance of this technique called FEAD (FEA based on subnetwork time distribution) is discussed by means of examples.<>