DDoS Attack Detection through Flow Analysis and Traffic Modeling

DDoS attack is the formidable cyber warfare of 20th century. Lot of research has already been taking place to mitigate DDoS attack. However DDoS attack still remains a potential threat. This research work considers the model level solution. Having a proper model of the traffic flow will help the administration unit to closely monitor the unusual behavior of the traffic; it will also help to identify the flash crowd which is the occasional accumulation of legitimate traffic. Hence in this paper, the normal traffic behavior is modeled, with the help of that the abnormal traffic which is evident during the DDoS attack is detected. Then the methodology to do the flow specific detection to segregate attack flow from the normal flow is discussed. Finally the possibility to curb the attack from the various hops is discussed.