Predictive Analytics as a Security Management Tool in Virtualised Environment

Despite the many benefits of virtualisation, security remains the main challenge for virtualisation in cloud computing. Even though a number of solutions like intrusion detection, malware introspection among others, have been developed and implemented by many security professionals, monitoring, visibility and infrastructure still pose security threats to virtualisation in cloud computing. In this paper, we propose the use of empirical predictive analytics as a security management tool in virtualised environments, as a novel contribution to security for virtualization in cloud computing. The proposed predictive analytics model is aimed at monitoring and analysing security threats unique to virtualisation in real time, providing us with the ability to mitigate these security threats before they actually happen. Predictive analytics, despite being an information technology process, has been successfully adopted in various disciplines like medicine, but not so much in information technology itself. The aim of this research is to be able to use empirical predictive analytics as a security management tool within a fully virtualised information system. In order for us to achieve empirical predictive analytics in a virtualised environment, a combination of concepts, tools and techniques, ranging from monitoring, machine learning, situation awareness, regression and instance-based learning techniques, among others, will be used. A brief description of what predictive analytics is and how it can be adopted as a security management tool in virtualised cloud computing environments is highlighted in the chapters that follow. The paper also gives a brief description of the proposed framework, the aims and objectives of the research and the methodology that will be used to achieve the objectives of the project.