A Machine Learning Approach for DDoS Detection on IoT Devices
This addresses DDoS detection for IoT security, but it is incremental as it applies existing methods to a new dataset.
The paper tackled the problem of detecting DDoS attacks on IoT devices by developing a machine learning model using the CICDDoS2019 dataset, achieving 100% accuracy in predicting network traffic types with AdaBoost and XGBoost algorithms.
In the current world, the Internet is being used almost everywhere. With the rise of IoT technology, which is one of the most used technologies, billions of IoT devices are interconnected over the Internet. However, DoS/DDoS attacks are the most frequent and perilous threat to this growing technology. New types of DDoS attacks are highly advanced and complicated, and it is almost impossible to detect or mitigate by the existing intrusion detection systems and traditional methods. Fortunately, Big Data, Data mining, and Machine Learning technologies make it possible to detect DDoS traffic effectively. This paper suggests a DDoS detection model based on data mining and machine learning techniques. For writing this paper, the latest available Dataset, CICDDoS2019, experimented with the most popular machine learning algorithms and specified the most correlated features with predicted classes are being used. It is discovered that AdaBoost and XGBoost were extraordinarily accurate and correctly predicted the type of network traffic with 100% accuracy. Future research can be extended by enhancing the model for multiclassification of different DDoS attack types and testing hybrid algorithms and newer datasets on this model.