EDoS-BARRICADE: A Cloud-Centric Approach to Detect, Segregate and Mitigate EDoS Attacks

Cloud computing through realizing the height of virtualization offers service models that can meet dynamic demands through performing auto-scaling of resources [1]. This helps the cloud service providers to broaden the grasp across sectors and the computing service market. Though it follows stretchable and elastic service models, it implements a rigid pay-per-use utility pricing model [Ribin Jones and Kumar in J Adv Res Dyn Control Syst 11(9):541–553, 2019 2]. The idea of dynamically scaling across platform makes it more vulnerable to security threats and makes room for easy exploits [Ribin Jones and Kumar in J Adv Res Dyn Control Syst 11(9):541–553, 2019 2]. Among various security threats, the economic denial-of-service (EDoS) attack presents a serious threat, since it exploits auto-scaling feature to impact the utility pricing model [Ribin Jones and Kumar in IEEE Xplore third international conference on trends in electronics and informatics, pp 1003–1008, 2019 3]. In this paper, a real-time cost incurring EDoS attack is performed against a cloud data center hosted Web page with simple Structured Query Language (SQL) manipulation method for experimental research. The experimental observations are applied to define an effective EDoS-BARRICADE that performs detection, segregation and mitigation specific to EDoS attack. The detection algorithm considers metrics that are associated with the auto-scaling feature to detect a suspicious increase in VM activities. The segregation algorithm implements linear SVM to isolate attack VMs optimally and rapidly. The results show that the developed EDoS-BARRICADE algorithms perform detection and segregation with 100% accuracy.