AESOP home

Publications

A functional central limit theorem for PEPA

Richard Hayden, Jeremy T. Bradley

National Workshop Paper
8th Workshop on Process Algebra and Stochastically Timed Activities
July, 2009
pp.13–23
Abstract

We present a functional central limit theorem which quantifies, as a stochastic process, the difference between a PEPA model's underlying CTMC and its fluid approximation. We generalise existing theory to handle the case of non-smooth rate functions, which is an issue particular to modelling problems in computer science. We verify the weak convergence empirically and suggest future avenues for deducing more analytic approximations from it.

PDF of full publication (503.5 kilobytes)
(need help viewing PDF files?)
PDF of presentation slides (1.1 megabytes)

Information from pubs.doc.ic.ac.uk/pepa-functional-central-limit.