Detecting Rapid Reset Attacks on 5G-Powered Industrial IoT using Machine Learning

5G networks have been widely used in modern-day networking, including its integration in the field of industrial IoT, serving as its backbone critical infrastructure, providing the communication that industrial systems require. However, this combination is vulnerable to various attacks, with rapid reset(RR) attacks being one of the most common. This paper utilizes an RR-based dataset that is part of the 5G flow dataset from the IEEE dataport to find a suitable ML algorithm to detect rapid reset attacks in 5G networks. The J48 variant of the decision tree scored an accuracy of 97.6524%, with a training time of 11.11s and over 97% in all other metrics. Additionally, the J48 consolidated variant achieved 99.0915% with a training time of 47.02s. Both outperform other traditional machine learning algorithms. A use case of the proposed IDS in industrial IoT is presented.