Volume 14, No 2, 2017
Hybrid Feature Selection And Classification Method For Intrusion Detection In Mobile Ad Hoc Networks
M. LALLI , V PALANISAMY
Abstract
Mobile Ad-hoc network has turned into an energizing and essential innovation as of late in view of the fast expansion of wireless gadgets. Mobile adhoc networks are highly vulnerable against attacks because of the progressively topology changing of network, the absence of centralized point for monitoring and open medium. The different attacks against mobile nodes are Warm hole, Byzantine attack, Packet Dropping, Black hole and flooding so on. It is vital to look new architecture and networks to ensure the wireless networks and the application of mobile computing. Intrusion Detection System devices are reasonable for distinguishing these attacks. It examine the network exercises by method for review information and use examples of surely understood attacks or ordinary profile to recognize potential attacks. In this paper, a hybrid framework to predict the intruding nodes in the Mobile ad-hoc network routing protocols. This framework is composed of three stages. The first stage of the framework is used to reduce the number of features of KDD CUP dataset which consists of six types of Denial of Service attacks by Hybridizing the Information Gain and Relative Reduct then optimal or reduced dataset is obtained. In the second stage of the framework, an Ant Colony Optimization is used for generating the rule structure for the six types of Denail of Service attacks using reduced features. In the second stage, from the optimal dataset, Naïve Bayes classification is used to classify the features into two categories are known attacks and unknown attacks. And a rule structure is generated by Ant Colony Optimization for known as well as unknown attacks. And in the final stage, Artificial Neuro Fuzzy Inference System is used to predict the behavior of the nodes in the Mobile ad-hoc network.
Pages: 101-123
Keywords: Mobile Ad Hoc Network, Denial of Service Attacks, Information Gain, Relative Reduct, Ant Colony Optimization, Naïve Bayes Classification, Artificial Neuro Fuzzy Inference System.