TY - JOUR
T1 - Graphical modelling for simulation and formal analysis of wireless network protocols
AU - Fehnker, A.
AU - Fruth, M.
AU - McIver, A. K.
PY - 2009
Y1 - 2009
N2 - It is well-known that the performance of wireless protocols depends on the quality of the wireless links, which in turn is affected by the network topology. The aim of this paper is to investigate the use of probabilistic model checking in the analysis of performance of wireless protocols, using a probabilistic abstraction of wireless unreliability. Our main contributions are first, to show how to formalise wireless link unreliability via probabilistic behaviour derived from the current best analytic models [12], and second, to show how such formal models can be generated automatically from a graphical representation of the network, and analysed with the PRISM model checker. We also introduce CaVi, a graphical specification tool, which reduces the specification task to the design of the network layout, and provides a uniform design interface linking model checking with simulation. We illustrate our techniques with a randomised gossiping protocol.
AB - It is well-known that the performance of wireless protocols depends on the quality of the wireless links, which in turn is affected by the network topology. The aim of this paper is to investigate the use of probabilistic model checking in the analysis of performance of wireless protocols, using a probabilistic abstraction of wireless unreliability. Our main contributions are first, to show how to formalise wireless link unreliability via probabilistic behaviour derived from the current best analytic models [12], and second, to show how such formal models can be generated automatically from a graphical representation of the network, and analysed with the PRISM model checker. We also introduce CaVi, a graphical specification tool, which reduces the specification task to the design of the network layout, and provides a uniform design interface linking model checking with simulation. We illustrate our techniques with a randomised gossiping protocol.
UR - http://www.scopus.com/inward/record.url?scp=63449131941&partnerID=8YFLogxK
U2 - 10.1007/978-3-642-00867-2_1
DO - 10.1007/978-3-642-00867-2_1
M3 - Article
AN - SCOPUS:63449131941
SN - 0302-9743
VL - 5454 LNCS
SP - 1
EP - 24
JO - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
JF - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
ER -