DoS and DDoS Attacks on Internet of Things and Their Detection by Machine Learning Algorithms

Machine Learning (ML) algorithms play a crucial role in fortifying the security of Internet of Things (IoT) environments. In this study, we focus on several key ML algorithms, namely Random Forest, AdaBoost, Decision Trees, Naive Bayes, Logistic Regression, Support Vector Machines (SVM), and k-Nearest Neighbors (k-NN). These algorithms are evaluated within the unique context of IoT security, employing an original dataset meticulously crafted for this study. The dataset is designed to capture the intricacies of cyber threats in an IoT network, featuring attacks such as DDoS, HTTP Flood, SYN Flood, Port Scan, and UDP Flood. This original dataset serves as a foundation for the comprehensive evaluation of ML algorithms, allowing us to assess their effectiveness in identifying and mitigating diverse attack patterns targeting IoT devices. The algorithms are examined based on their performance metrics such as accuracy, F1-score, precision and recall, emphasizing their suitability for real-world IoT security applications. The results show that Random Forest and AdaBoost are the top performers in terms of performance metrics. The study aims to provide valuable insights into the strengths and limitations of these ML algorithms, aiding researchers and practitioners in developing resilient security measures designed for IoT settings.