Title: Extending GSPNs for the modelling, analysis and performance evaluation of dynamic systems

Authors: Samir Tigane; Laid Kahloul; Samir Bourekkache; Souheib Baarir

Addresses: LINFI Laboratory, Computer Science Department, Biskra University, Biskra, Algeria ' LINFI Laboratory, Computer Science Department, Biskra University, Biskra, Algeria ' LINFI Laboratory, Computer Science Department, Biskra University, Biskra, Algeria ' LIP6 Laboratory, Computer Science Department, Paris Ouest Nanterre la Defense University, Paris, France

Abstract: Cloud-based systems are open systems where resources are theoretically non-limited and the number of connected users/devices varies all the time. These systems are close to dynamic systems where the structure changes at runtime. The objective of this paper is to present a new formalism 'extended generalised stochastic Petri nets (EGSPNs)' dedicated to model, analyse and evaluate the performance of reconfigurable systems. In its basic form, GSPNs extend Petri nets by introducing stochastic time. In this paper, GSPNs are enriched with a set of mechanisms allowing the reconfigurability of their structure at runtime. The new formalism 'EGSPNs' is applied to model, analyse and evaluate the performance of a case study.

Keywords: Petri nets; generalised stochastic Petri nets; rewriting systems; extended GSPNs; reconfigurable systems; formal modelling; qualitative verification; performance evaluation; dynamic systems; critical systems.

DOI: 10.1504/IJCCBS.2018.091837

International Journal of Critical Computer-Based Systems, 2018 Vol.8 No.1, pp.25 - 44

Received: 16 Mar 2017
Accepted: 09 Dec 2017
Published online: 18 May 2018 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article