Response to Co-resident Threats in Cloud Computing Using Machine Learning

Abstract

Virtualization technologies in cloud computing brings merits for resource utilization and on-demand sharing. However, users can face new security risks when they use the virtualized platforms. The co-resident attack means that the malicious users build side channels and threaten the virtual machines co-located on the same server. In 2017, Abazari et al. proposed a multi-objective, under the constraints of minimum cost and threat, response to co-resident attacks in cloud environment. In this paper, we aimed to propose a novel method for countermeasures decision to the attacks. We used machine learning to train the intrusion response model and conducted a set of experiments to demonstrate the effect of the proposed model. It showed that the near optimal solutions with good accuracy is obtainable and the response efficiency is improved.