Detecting wormhole attack in Environmental Monitoring System for Agriculture using Deep Learning

The Internet of Things (IoT) is a rapidly growing field that connects various devices and systems to the internet, enabling them to communicate and share data. However, this increased connectivity also makes IoT networks vulnerable to various types of attacks, one of which is the wormhole attack. A wormhole attack is a type of security threat in which an attacker creates a tunnel between two or more nodes in an IoT network, allowing the attacker to intercept, modify or inject malicious packets into the network. This can lead to serious security issues such as unauthorized access, data leakage and network disruption. The problem of wormhole attack detection in IoT networks is a crucial issue that must be addressed. Traditional security methods, such as firewalls and intrusion detection systems, may not be effective in detecting and preventing wormhole attacks, as these attacks are difficult to detect due to the stealthy nature of the attacker. Therefore, there is a need for new and more advanced methods for wormhole attack detection in IoT networks, such as deep learning approaches. The goal of this paper is to use a deep learning approach to detect wormhole attacks in IoT networks and to compare the performance of this approach with traditional machine learning methods. This research paper presents a deep learning approach for wormhole attack detection in Internet of Things (IoT) networks using Long Short-Term Memory (LSTM) model. The proposed method is compared with traditional machine learning techniques which are Decision Tree, and Naive Bayes. The performance of the proposed approach is evaluated using a malware dataset for predicting the type of wormhole attack (WHR). The evaluation metrics used in this study include accuracy, F1 score, precision, recall and confusion matrix. The implementation of the proposed approach is performed using Python programming and the Anaconda Navigator (Spyder notebook) tool. The results show that the proposed LSTM-based approach outperforms traditional machine learning techniques in terms of accuracy and F1 score which is 99% while Decision Tree Model accuracy is 94% and Naïve Bayes Model scores 93%, the output results of this paper demonstrating the effectiveness of deep learning in wormhole attack detection in IoT networks.

Text 120641.pdf - Published Version Available under License Creative Commons Attribution Non-commercial. Download (6MB) Official URL or Download Paper: https://semarakilmu.com.my/journals/index.php/appl...

Actions (login required)

View Item