WNetMon: An ML Approach for Real-time DoS Attack Detection in Wireless Networks

Denial of Service (DoS) attacks have emerged as sophisticated threats that exploit the known vulnerabilities of wireless communication, potentially sabotaging their operations and causing extensive downtime. Among these, de-authentication, disassociation and beacon flooding are particularly concerning due to their efficiency in disrupting network services. This paper delves into different types of Denial of Service (DoS) attacks and proposes a state-of-the-art detection mechanism. Most of the current state-of-the-art ML and DL-based IDSs are evaluated on the training dataset, which needs more variation in real-time attack data and requires substantial computational resources. In the following work, we propose a lightweight software solution named WNetMon, developed using the AWID2 dataset and evaluated using a new dataset generated using our custom testbed. Moreover, it can perform real-time Denial of Service (DoS) flooding attack detection on edge devices in wireless networks. Our results show that, while being small and effective, WNetMon achieves an overall accuracy of 99% for attack detection, benchmarked in real-time network traffic generated in our testbed. Therefore, it demonstrates the potential for using extensible ML solutions for Denial of Service (DoS) attack detection on edge systems that cannot execute industrial network monitoring tools due to resource constraints.