Citation

Halbouni, Asmaa and Ong, Lee Yeng and Leow, Meng Chew (2025) From Traditional Signatures to Deep Learning: Enhancing WPA3 Security Against Deauthentication Attacks. IEEE Access, 13. pp. 182796-182808. ISSN 2169-3536

Text 1.pdf - Published Version Restricted to Repository staff only Download (4MB)

Abstract

Intrusion Detection Systems play an important role in protecting against cyberattacks. Traditional signature-based IDS detect known attacks using rules and signatures, deep learning-based IDS can detect known and unknown attacks through analyzing their patterns. WPA3, the latest security protocol, although outperformed its predecessors but still suffers from some attacks such as Deauthentication attacks. This research examines the effectiveness of traditional Intrusion Detection System, namely Snort, against deep learning Intrusion Detection System, namely CNN-DNN in detecting deauthentication attacks in WPA3 wireless networks. Such a combination of CNN and DNN will make the model more reliable in detecting the attack. To achieve the goal in this research, a dataset has been collected through Kismet software comprising normal traffic and deauthentication traffic in WPA3 wireless network. The collected dataset composed of normal traffic and deauthentication attack traffic based on WPA3 wireless network based on controlled environment. The collected dataset has been used then by Snort model and CNN-DNN model to test their abilities in detecting deauthentication attack. The confusion matrix is used to evaluate the performance of both models. Based on the experimental results obtained, CNN-DNN improved performance and provided more reliability in detecting deauthentication attack traffic. The results obtained in terms of accuracy, precision, detection rate, and false alarm rate for Snort model and CNN-DNN model are: (67% and 95.67%), (64% and 92.7%), (40% and 89.36%), and (15% and 22%), respectively. To add more credibility to this research, a significance test has been made to confirm that CNN-DNN model outperformed Snort model.

Item Type:	Article
Uncontrolled Keywords:	convolutional neural network, intrusion detection system, Snort, WPA3 protocol
Subjects:	Q Science > QA Mathematics > QA71-90 Instruments and machines > QA75.5-76.95 Electronic computers. Computer science
Divisions:	Faculty of Information Science and Technology (FIST)
Depositing User:	Nor Afiqah Mohd Adnan
Date Deposited:	10 Dec 2025 03:03
Last Modified:	13 Dec 2025 03:40
URII:	http://shdl.mmu.edu.my/id/eprint/15015

Downloads

Edit (login required)