Enhancing IoT Network Security: A ML-Based Intrusion Detection System

The Internet of Things (IoT) consists of a network of devices that connect and interact over the Internet, allowing them to share data seamlessly. These devices include everything from everyday household appliances to complex industrial machinery, enhancing efficiency, convenience, and productivity. However, as IoT networks grow in popularity, they become more vulnerable to security breaches. Cyberattacks pose significant threats to IoT security, making it crucial to detect these attacks to protect IoT networks. Machine learning (ML) techniques have been employed as intrusion detection systems (IDSs) to enhance security capabilities. To support this effort, datasets like TON-IOT have been developed. In our study, we use the TON-IOT dataset, created at the UNSW Canberra lab, which is the largest dataset available for detecting cyberattacks on IoT networks. This dataset includes data from various IoT sensors, such as thermostats, refrigerators, garage doors, GPS trackers, motion lights, and weather sensors. Using the Random Forest, XGBoost and Light Gradient Boosting Machine (LGBM), our goal is to find the best machine learning approach to achieve the highest possible accuracy and identify a better method compared to previous research in this area, machine learning(ML).