Trust Aware Svm Based Ids For Mitigating The Malicious Nodes In Manet
A.R.Rajeswari1, K.Kulothungan2, S.Ganapathy3, A.Kannan4
1A.R.Rajeswari, Department of Computer Science and Engineering, Sethu Institute of Technology, kariapatti, India
2K.Kulothungan, Department of Information Science and Technology, CEG Campus, Anna University, Chennai, India
3S.Ganapathy, School of Computing Science and Engineering, VIT University-Chennai Campus, Chennai, India
4A.Kannan, Department of Information Science and Technology, CEG Campus, Anna University, Chennai, India
Manuscript received on 02 June 2019 | Revised Manuscript received on 10 June 2019 | Manuscript published on 30 June 2019 | PP: 185-197 | Volume-8 Issue-8, June 2019 | Retrieval Number: G6278058719/19©BEIESP
Open Access | Ethics and Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: MANET is dynamic in nature, openness, infraturestureless and no centralized monitoring and controlling unit. Due to these unique characteristic features, MANET is subjected to various security threats caused by the malicious nodes. A system that observes the unwanted attacks caused by the malicious nodes is defined as Intrusion Detection System. The major responsibility of IDS is to detect attacks from the network. In this paper, we propose a Trust Aware SVM based Intrusion Detection System (TASVM-IDS) with an objective to detect and isolate the malicious nodes. The proposed system consists of the following modules, namely feature extraction module, trust estimation module, classification module and decision making model. In this paper, a novel feature extraction algorithm, namely Linear Correlation Coefficient Based Feature Extraction (LCCBFE) algorithm is proposed with an aim to minimize the training time and to enhance the lifetime of the system. The trust level node is estimated by utilizing the behavior analysis and residual energy level of nodes. Thus, we have proposed a new Behavior Analysis Based Trust Algorithm (BABT) algorithm to compute the trust level of nodes in the network. Finally, SVM based classifier is used to classify the nodes into a trustworthy, untrustworthy or malicious node based upon the measured trust level of the nodes. Simulation results proves that the proposed TASVM-IDS can successfully mitigate malicious node and gives better results when compared to SVM and ELM.
Keyword: Detection System; Data preprocessing; SVM; ELM; Trust; MANET.
Scope of the Article: Agent-Based Software Engineering