Abstract: The wormhole attack is a type of attack on the network layer that affects the routing protocols. In this study, we investigate the use of Gated Recurrent Unit (GRU) with an attention mechanism and Decision Tree, for detecting wormhole attacks in mobile ad hoc networks (MANETs). This study is divided into three main tasks. Firstly, simulating wormhole attacks in a mobile ad hoc networks (MANETs) environment with a finite number of nodes Secondly, describe the network characteristics that contribute to feature selection. Consequently, we perform data generation and data gathering operations that produce a large volume of datasets. Finally, we applied the proposed model to detect wormhole attacks in mobile ad hoc networks (MANETs). The model was evaluated and tested with the simulated datasets consisting of eight selected features and an instance using Accuracy, Precision, Recall, and F1-Score as performance metrics. Experimental results demonstrate that the proposed method outperformed other related methods from the literature in terms of the aforementioned evaluation criteria.
Keywords: Manets, Wormhole, Attention mechanism, DT, Feature Selection.
Title: Attention Based Gated Recurrent Neural Network for Wormhole Attack Detection in MANETs
Author: Balkisu Musa Hari, Ali Ahmad Aminu
International Journal of Computer Science and Information Technology Research
ISSN 2348-1196 (print), ISSN 2348-120X (online)
Vol. 11, Issue 4, October 2023 - December 2023
Page No: 20-28
Research Publish Journals
Website: www.researchpublish.com
Published Date: 10-October-2023
