AESOP home

Publications

Semi-Markov PEPA: Modelling with Generally Distributed Actions

Jeremy T. Bradley

Journal Article
International Journal of Simulation
Volume 6
Issues 3–4
pp.43–51
February, 2005
IJSIM
ISSN 1473-804X
Abstract

Since the advent of Markovian Process Algebras, users have requested the ability to employ a greater variety of action distribution. We present a conservative extension to the popular Markovian process algebra, PEPA, which incorporates generally distributed sojourn-times for action duration. Just as a PEPA model generates a Markov chain for analysis purposes, so semi-Markov PEPA produces a semi-Markov chain. We discuss how semi-Markov PEPA models are analysed through Dingle and Knottenbelt's semi-Markov DNAmaca tool and present a small example for passage time analysis.

PDF of full publication (135.8 kilobytes)
(need help viewing PDF files?)
GZipped Postscript of full publication (552.8 kilobytes)
(need help viewing GZipped Postscript files?)

Information from pubs.doc.ic.ac.uk/semi-markov-pepa.