A novel statistical technique for detection of DDoS attacks in KDD dataset

Recent times have seen a surge in the number of Distributed Denial-of-Service (DDoS) attacks as the attackers incessantly come up with new and sophisticated techniques to carry out such attacks. The most robust solution to this problem in many cases is a fundamental method that looks for anomalies in network traffic. These anomalies are further classified into attacks and their variations using certain techniques and parameters. In this paper, statistical technique of detecting DDoS attacks by observing network traffic and looking for anomalous behavior in it has been described. For the experiment, the dataset used in the 3rd International Knowledge Discovery and Data Mining Tools Competition (1999) has been used. The dataset is commonly called the KDD dataset. Using the proposed technique, it is possible to detect a number of DDoS attacks and also tell the approximate time of their occurrence.