Enhancing Anomaly Based Intrusion Detection Techniques for Virtualization in Cloud Computing Using Machine Learning

Abstract

Cloud computing is a rapidly growing technology that enables elastic, easy to manage, cost effective and on demand access to powerful computational resources for instance applications storage, servers, networks and services on internet. Convenient and ubiquitous deployment of clouds make it appealing to the users. The core component of cloud computing is virtualization, that is used to create multiple virtual machines from single physical machine in less cost. These Vms have a significant importance because of their ineludible utilization. Despite of all vital advantages, cloud computing domain is inclined to numerous security concerns. Intrusion is one such obstacle that can affect integrity and confidentiality of data. Host based intrusion detection systems have validated to be important in identifying network traffic and performance metrics of the guest machine maintained by hypervisor which do not conform to authorized patterns. In the current research, the aim is to make use of anomaly-based IDS to observe and recognize activities that are not part of legitimate network traffic in cloud computing environment using virtual machines. The proposed methodology observes the traffic of Virtual Machines. Scan and detect the anomalous behavior and in response restrain network traffic with anomalous packets.