Enhanced-Adaptive Pattern Attack Recognition Technique (E-APART) Against EDoS Attacks in Cloud Computing

Abstract

Cloud Computing is most widely used in current technology. It provides a higher availability of resources to greater number of end users. In the cloud era, security has develop a reformed source of worries. Distributed Denial of Service (DDoS) and Economical Denial of Sustainability (EDoS) are attacks that can affect the ‘pay-per-use' model. This model automatically scales the resources according to the demand of consumers. The functionality of this model is to mitigate the EDoS attack by some tactical attacker/s, group of attackers or zombie machine network (BOTNET) to minimize the availability of the target resources, which directly or indirectly reduces the profits and increase the cost for the cloud operators. This paper presents a model called Enhanced-APART which is step further of the authors' previous model (APART) that can be used to mitigate the EDoS attack from the cloud platform and shows the nature of the attack. Enhanced-APART model offers pre-shared security mechanism to ensure the access of legitimate users on the cloud services. It also performs pattern analysis in order to detect the EDoS caused by BOTNET mechanism and includes time-based and key-sharing post-setup authentication scheme to prevent the replication or replay attacks and thus results in mitigation of EDoS attack.