A Hybrid Approach for IEEE 802.11 Intrusion Detection Based on AIS, MAS and Naïve Bayes

Abstract

The advancement of network technology has been offering substantial improvements for users in general. Paradoxically, new technologies end up bringing also new types of problems. Not only one can find issues related to architecture failure, mis-configuration and bad adaptation, but also a growing number of problems related to user security. The latter is a problem strongly connected to wireless networks. Obviously, this is facilitated by the means of transport used in those networks (radio waves). Intrusion attempts are common and a strong concern regards detecting them. To this end, due to the fact that it is easy to attack and tough to defend wireless networks, good new approaches would be the ones that could profit from intelligent techniques as they are adaptive and thus may identify attacks that are not necessarily known in advance. In this work we use the Danger Theory (DT) and a Bayesian classifier (naïve Bayes) embedded into a military style multi-agent (MAS) to create a light, dynamic and adaptive detection system to work with the lower layer of wireless networks (WIDS). Experimental results show that the artificial immune aspect of the system is capable of detecting unknown issues and to identify them automatically with considerable few false alarms and low cost for the network traffic.