Title: A hidden Markov model based intrusion detection system for wireless sensor networks

Authors: L. Coppolino; L. Romano; A. Bondavalli; A. Daidone

Addresses: Dipartimento per le Tecnologie, University of Naples Parthenope, Centro Direzionale di Napoli Is. C4, 80143 Naples, Italy. ' Dipartimento per le Tecnologie, University of Naples Parthenope, Centro Direzionale di Napoli Is. C4, 80143 Naples, Italy. ' Dipartimento di Sistemi e Informatica, University of Firenze, Viale Morgagni 65, 50134 Firenze, Italy. ' Resiltech srl, piazza Nilde Iotti, 25 56025 Pontedera (PI), Italy

Abstract: Wireless sensor network (WSN) technology is being increasingly used for data collection in critical infrastructures (CIs). This paper presents an intrusion detection system (IDS), which is able to protect a CI from attacks directed to its WSN-based parts. By providing accurate and timely detection of malicious activities, the proposed IDS solution ultimately results in a dramatic improvement in terms of protection, since opportunities are given for performing proper remediation/reconfiguration actions, which counter the attack and/or allow the system to tolerate it. The proposed solution has the important advantage of exploiting the high accuracy of hidden Markov models as an effective means of detecting malicious activities. We present the basic ideas, discuss the main implementation issues, and perform a preliminary experimental campaign, with respect to sinkhole attacks, one of the most serious attacks to WSNs.

Keywords: critical infrastructures; infrastructure protection; intrusion detection; intrusion diagnosis; hidden Markov models; HMMs; wireless sensor networks; WSNs; wireless networks; malicious activities; sinkhole attacks; remediation; reconfiguration.

DOI: 10.1504/IJCCBS.2012.050305

International Journal of Critical Computer-Based Systems, 2012 Vol.3 No.3, pp.210 - 228

Received: 06 Feb 2012
Accepted: 10 Jul 2012
Published online: 16 Aug 2014 *

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