Ensemble Classifiers with Drift Detection (ECDD) in Traffic Flow Streams to Detect DDOS Attacks

Malfunction of internet networking systems might directly and the adverse effect in one way or the other, wherein aspects of contemporary information and communication technologies. In such conditions, DDoS attacks are prevalent threat, wherein flooding of requests related to computation and communication resources for ordering the service that is unavailable for legitimate users. DDOS attacks to be defend to guard the Critical resources. The contribution of this manuscript is an ensemble classifier model to defend the DDOS attacks. The Proposed model is based on ensemble classifier with drift detection ability at the service request stream level. The proposed model incorporates the process of defining service request streaming characteristics, enables the drift detection ability that uses the defined service request stream characteristics. The experimental study carried out from the setup established using synthesized service request stream, and the result obtained are explored using statistical metrics such as true negative rate, positive predictive value, accuracy. Moreover, the significance of the model elevated by comparing the obtained results with results obtained from other benchmark models depicted in contemporary literature.