Title: Modelling misbehaviour in ad hoc networks: a game theoretic approach for intrusion detection

Authors: Yu Liu, Cristina Comaniciu, Hong Man

Addresses: Department of Electrical and Computer Engineering, Stevens Institute of Technology, Hoboken, NJ, USA. ' Department of Electrical and Computer Engineering, Stevens Institute of Technology, Hoboken, NJ, USA. ' Department of Electrical and Computer Engineering, Stevens Institute of Technology, Hoboken, NJ, USA

Abstract: In wireless ad hoc networks, although defence strategies such as Intrusion Detection Systems (IDSs) can be deployed at individual nodes, significant constraints are imposed in terms of the energy expenditure of such systems. In this paper, we use a game theoretic framework to analyse the interactions between pairs of attacking/defending nodes, in both static and dynamic contexts, and considering both complete and incomplete information regarding the maliciousness of neighbouring nodes. The static case analysis provides the defender with an overview of the security situation in terms of risk and monitoring cost. A dynamic Bayesian game formulation allows the defender to adjust his belief about his opponent based on his observations and the game history, and consequently to influence the achievable Nash equilibrium for new stage game. A new Bayesian hybrid detection system is proposed for the defender, which balances energy costs and monitoring gains.

Keywords: ad hoc networks; non-cooperative games; non-zero-sum games; Bayesian games; attacker-defender games; wireless networks; security; hybrid detection; game theory; intrusion detection; risk costs; monitoring costs; energy costs; misbehaviour.

DOI: 10.1504/IJSN.2006.011784

International Journal of Security and Networks, 2006 Vol.1 No.3/4, pp.243 - 254

Published online: 21 Dec 2006 *

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