17. Advanced Machine Language Approach to Detect DDoS Attack Using DBSCAN Clustering Technology with Entropy

Service availability is the major and primary security issue in cloud computing environments. Currently existing solutions that address service availability-related issues that can be applied in cloud computing environments are insufficient. In order to ensure the high availability of the offered services, the data centers resources must be protected from DDoS attack threats. DDoS is the major and the most serious security threat that challenges the availability of the data centers resources to the intended clients. The existing solutions that monitor incoming traffic and detect DDoS attacks become ineffective if the attacker’s traffic intensity is high. Therefore, it is necessary to devise schemes that will detect DDoS attacks even when the traffic intensity is high; such schemes must deactivate DDoS attackers and serve the legitimate users with available re-sources. This research paper addresses the need to prevent DDoS attacks by defining and demonstrating a hybrid detection model by introducing an advanced and efficient approach to recognize and efficiently discriminate the flood attacks from the flush crowd (legitimate access). Moreover, this paper introduce and discusses, most importantly, the application of multi-variate correlation among the selected and ranked features to significantly reduce the false alarm rate, which is one of the major issue associated with the current available solution.