DDoS attack detection and classification via Convolutional Neural Network (CNN)

Abstract

Distributed Denial of Service (DDoS) attacks became the most widely spread attack because it is easily designed and executed but it is very difficult to detect and mitigate. Several artificial neural network (ANN) techniques were considered to detect and classify DDoS attacks. Mission control center (MCC) is responsible for controlling the spacecraft, so MCC network should maintain the availability i.e. should be protected from any kind of malicious traffic affect its availability such as DDoS attack. In this paper, convolutional neural network (CNN) technique is presented to detect and classify the DDoS traffic into normal and malicious information with an accuracy of 99 % using two different datasets. One is captured from simulated MCC network by Wireshark and the other one was a predefined open source dataset. The results are compared with other classification algorithms like decision tree (D-Tree), support vector machine (SVM), K-nearest neighbors (K-NN), and neural network (NN).