Abstract

Distributed denial of service (DDoS) attacks is one of the most destructive cyber attacks which target the availability of the system when legitimate users try to access the system. Not only computers, but also the growing number of smartphones as well as Internet of Things (IoT) devices are affected by DDoS attacks. There is no well-known system which effectively stops or prevents DDoS attacks. Designing an effective DDoS detector with high accuracy with low computational overhead is still a very challenging task. In this paper, a methodology, which is used to detect and classify the types of DDoS attacks, is proposed. Our methodology is divided into three parts: pre-processing, feature selection, and classification. First, pre-processing is performed to eliminate some features which are not suitable for our model. Second, most significant features are selected by using Information Gain, Gain Ratio, Correlation Coefficient, and Relief. We declined the number of features from 87 to 20. Finally, various classifiers are used to detect DDoS attacks from the bening ones. The proposed methodology is performed on the CIC-DDoS2019 dataset. The experimental results show that the proposed methodology performed pretty well when it is compared to leading methods in the literature.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call