Detection of DoS Attacks in Internet of Things Networks: A Comparative Study
Mohamad Abdallah Mohamad, Mohamed Wed Eladham, Mohamed Saad · 2024
In the era of the Internet of Things (IoT), securing IoT devices is crucial due to the dual nature of enabling seamless information flow and exposing vulnerabilities to cyber threats. This research focuses on identifying Denial-of-Service (DoS) attacks within IoT environments, utilizing Machine Learning techniques. This work significantly contributes to enhancing IoT security by strengthening detection and defence mechanisms against DoS attacks. With the rising integration of diverse devices into IoT networks, protection against security threats becomes necessary. Assessing five machine learning methods on the Bot-IoT dataset, the study reveals Gradient Boosting as the top performer, classifying DoS attacks with 99.19% accuracy. Future research could extend to detecting various attack types and conduct a comparative analysis between deep learning and classical machine learning models for effective intrusion detection on IoT networks.