Novel Intrusion Detection in MANETs based on Trust

Mobile Ad hoc Network is comprised of several mobile nodes without any centralized infrastructure. MANETs are lack of boundaries due to wireless communication medium. Due to the absence of centralized infrastructure, changing topology and distributed nature, MANETs is prone to several security threats and intruders. Thus, it is necessary to safeguard the network from malicious nodes. Thus to overcome with this security issue we need to develop a robust, flexible intrusion detection system. This paper aims at detecting the malicious nodes by proposing an Intrusion Detection System (IDS). We plan to employ KDD’99 dataset and the features are planned to get selected by mutual information and information gain ratio. A trust based multiclass Extreme Learning Machine (ELM) is planned to be incorporated for effective classification of malicious nodes. The expected outcome of the system will be effective as the system considers trust value of nodes. The performance metrics that will be used to check the performance of the system are misclassification rate, detection accuracy, execution time and false alarm rate under different scenarios. Thus our work provides novel way of Intrusion Detection. Keywords— Intrusion Detection System, MANETs, Trust, Anomaly Detection, Malicious node.