Aggregation methods in exact performance analysis of stochastic Petri nets

Abstract

Recently the authors have shown that there is a class of stochastic Petri nets (SPNs), distinct from queuing networks, that also have a product-form equilibrium distribution. In this study the previous work is extended by study in SPNs that have arbitrarily distributed enabling and firing times, probabilistic input and output bags, and colored tokens. An embedding procedure that reduces a complex SPN to a skeleton SPN is proposed. It is shown that choosing the appropriate embedding, it is possible to obtain exact results from the skeleton SPN that are marginal results for the complex SPN.