Detection and Mitigation of DDoS Attacks on IoT Networks Using Machine Learnin
Akram Rashid, Syed Muhammad Khaliq-ur-Rahman Raazi, Shayhaq Ali · 2024
Smart homes, cities, and healthcare are just a few examples of how the Internet of Things (IoT) simplifies daily life. With the rise of 5G, the number of connected IoT devices is rapidly increasing, expected to surpass 75 billion by 2025. However, this growth also brings significant security risks, particularly the vulnerability of IoT Network to Distributed Denial of Service (DDoS) attacks, which can overwhelm networks. In this research, we applied machine learning techniques on the CIC-DDoS 2019 dataset to detect and mitigate DDoS attacks on IoT devices. The results show high accuracy across several models: Logistic Regression (88%), Decision Tree (93%), Random Forest (94%), KNN (93%), Naive Bayes (72%), and SVM (89%). These findings suggest that machine learning is an effective approach for real-time DDoS attack detection and mitigation in IoT networks.