Effect of Clustering in Designing a Fuzzy Based Hybrid Intrusion Detection System for Mobile Ad Hoc Networks
D. Vydeki and R. S. Bhuvaneswaran
DOI : 10.3844/jcssp.2013.521.525
Journal of Computer Science
Volume 9, Issue 4
Intrusion Detection System (IDS) provides additional security for the most vulnerable Mobile Adhoc Networks (MANET). Use of Fuzzy Inference System (FIS) in the design of IDS is proven to be efficient in detecting routing attacks in MANETs. Clustering is a vital means in the detection process of FIS based hybrid IDS. This study describes the design of such a system to detect black hole attack in MANET that uses Adhoc On-Demand Distance Vector (AODV) routing protocol. It analyses the effect of two clustering algorithms and also prescribes the suitable clustering algorithm for the above-mentioned IDS. MANETs with various traffic scenarios were simulated and the data set required for the IDS is extracted. A hybrid IDS is designed using Sugeno type-2 FIS to detect black hole attack. From the experimental results, it is derived that the subtractive clustering algorithm produces 97% efficient detection while FCM offers 91%. It has been found that the subtractive clustering algorithm is more fit and efficient than the Fuzzy C-Means clustering (FCM) for the FIS based detection system.
© 2013 D. Vydeki and R. S. Bhuvaneswaran. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.