Author(s)

Yun Feng, Fu Kun

Corresponding Author

Yun Feng

ABSTRACT

A growing number of Internet of Things (IoT) devices are appearing on the Internet. Yet these devices are facing more and more security risks, such as Distributed denial of service (DDoS) attack. In recent years, the scale of DDoS attacks is getting larger and larger, and each DDoS attack will cause huge losses. This motivates the development of new techniques to automatically detect attack traffic and improve detection accuracy as much as possible, so as to reduce losses. This paper demonstrates several ensemble learning methods which show a higher accuracy DDoS detection in IoT network traffic than using a single machine learning method. Experimental results show that ensemble learning can complement the limitations of using a single machine learning method to improve the DDoS detection accuracy.

KEYWORDS

Internet of Things, Anomaly Detection, DDoS, Machine Learning, Feature Engineering, Ensemble Learning