In the present days, mobile ad hoc networks (MANET) become popular because of the self-configuring or autonomous nature of nodes. However, wireless communication and node distribution in MANET makes it easily susceptible to malicious attacks. Consequently, it is needed to design effective intrusion detection systems (IDS) to defend the MANET from various kinds of attacks. In this paper, we employ deep learning (DL) based classification model for intrusion detection in MANET. Since the classification accuracy is mainly based on the nature of features that exist in the data, feature selection approaches are preferred to eliminate unwanted and repetitive data. Keeping this in mind, in this paper, whale optimization algorithm (WOA) is employed to select features and thereby enhances the classification accuracy of the deep learning model. The presented model is tested against the benchmark KDD Cup 99 dataset. Comparing with the contemporary methods, the introduced model demonstrated superior performance interms of various measures.
Volume 11 | 10-Special Issue
Pages: 1374-1387
DOI: 10.5373/JARDCS/V11SP10/20192983