Detection and Analysis of Cyber-Attacks on IoT Network Devices

Abstract

One of the most pressing concerns in network forensics is the detection of cyber-attacks in the Internet of Things (IoT) networks and their devices. Traditional intrusion detection systems based on signature rules are unable to detect current attack types. Hence, the need to urgently develop advanced methods for classifying IoT network traffic that can swiftly detect cyber-attacks becomes inevitable. This research aims to develop machine learning algorithms for cyber-attack detection in IoT-based networks, by analyzing the traffic data composed from the network itself. An ideal IoT network was implemented solely for the attack scenarios and generation of a dataset. For this study, both the structure and security of IoT networks were investigated in detail, by utilizing an IoT network created in real environment. Attack scenarios have been created for IoT devices in the real environment created. An IoT security data set was created by collecting the network flows obtained as a result of the attacks. On the created data set, classification results close to 100% accuracy values were obtained with machine learning algorithms. The data set obtained in the study has been published publicly so that other researchers can use it. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.