<![CDATA[The Internet of Things (IoT) is a technological innovation that enables physical devices to connect and communicate with one another through the internet, allowing them to exchange data automatically. This technology has been widely applied in various domains, such as smart homes and security systems. However, as the number of connected devices increases, the risk of cyberattacks also grows. One common type of attack is Denial of Service (DoS), an attempt to flood a system with excessive traffic, thereby disrupting communication between devices. This attack often exploits the MQTT protocol, which is popular in IoT environments due to its lightweight and efficient nature. This study aims to detect DoS attacks in MQTT-based IoT systems by implementing the Extreme Gradient Boosting (XGBoost) algorithm combined with feature selection using Particle Swarm Optimization (PSO). The dataset used consists of simulated MQTT traffic designed to resemble real-world conditions. The developed model is capable of classifying data into either normal or attack categories. The evaluation results demonstrate excellent performance, with precision ranging from 94.44% to 96.80%, recall from 99.63% to 99.97%, F1-scores between 96.95% and 97.81%, and an average accuracy of 99.89%. The main contribution of this research lies in integrating XGBoost with PSO on a realistic MQTT-based simulation dataset. This approach enhances both accuracy and computational efficiency, making it more suitable for resource-constrained IoT devices, and underscores the novelty of producing a DoS attack detection system that is accurate, efficient, and adaptive to real IoT network conditions.]]>
Copyrights © 2025
