An Efficient Anomaly Detection Framework for Cloud Computing Environment
Ming‐Wei Lin, Shuyu Chen · Journal of Computers · 2015
Infrastructure as a Service (IaaS) is an important service type provided by cloud computing.Infrastructure resources are encapsulated into services and they are provided to users over the Internet in the form of virtual machines.A malicious user can upload malicious software into the virtual machine allocated by a cloud computing service provider and launch the side channel attacks to other virtual machines located in the same physical node by operating his own virtual machine.In order to address the above problem, this paper proposes an efficient anomaly detection framework for cloud computing environment to detect the virtual machines that present abnormal behaviors.A new feature extraction algorithm is designed to reduce the dimensionality of the collected data and a new anomaly detection algorithm is also designed to detect the abnormal virtual machines.A series of experiments are conducted on a cloud computing environment that is deployed using the open source project OpenStack to evaluate the proposed framework.Experimental results show that the proposed framework is better than other anomaly detection methods designed for cloud computing environment in terms of precision, recall, false alarm rate, and runtime.