Revealing of Reducing Manners in Ad Hoc Networks with Crosslayer Approach Using SVM and FDA in Distributed Architecture

Abstract

Ad hoc network is a structure less network with independent nodes. In the ad hoc network, the nodes have to cooperate for services like routing and data forwarding. The routing attacks in ad hoc networks have given rise to the need for designing novel intrusion detection algorithms, different from those present in conventional networks. In this work, distributed intrusion detection system (IDS) have proposed for detecting malicious sinking behavior in ad hoc network. Detection process of that sinking behavior node is very important to do the further forwarding process in network. Intrusion detection system use linear classifiers for training the intrusion detection model. Cross -layer approach is involved to increase the accuracy of intrusion detection process in ad hoc network. A machine learning algorithm in non linear manner named as Support Vector Machine (SVM) involved for training the detection system and used together with Fisher Discriminant Analysis (FDA). The proposed cross-layer approach aided by a combination of SVM and FDA reduces the feature set of MAC layer without reducing information content.