Numerically Representing A Stochastic Process Algebra

Abstract: The syntactic nature and compositionality characteristic of stochastic process algebras make models to be easily understood by human beings, but not convenient for machines as well as people to directly carry out mathematical analysis and stochastic simulation. This paper presents a numerical representation schema for the stochastic process algebra PEPA, which can provide a platform to directly and conveniently employ a variety of computational approaches to both qualitatively and quantitatively analyse the models. Moreover, these approaches developed on the basis of the schema are demonstrated and discussed. In particular, algorithms for automatically deriving the schema from a general PEPA model and simulating the model based on the derived schema to derive performance measures are presented.