Security in Cloud Computing: Evaluation and Integration

Over the past decade, the Cloud Computing paradigm has revolutionized the way we envision IT services. It has provided an opportunity to respond to the ever increasing computing needs of the users by introducing the notion of service and data outsourcing. Cloud consumers usually have online and on-demand access to a large and distributed IT infrastructure providing a plethora of services. They can dynamically configure and scale the Cloud resources according to the requirements of their applications without becoming part of the Cloud infrastructure, which allows them to reduce their IT investment cost and achieve optimal resource utilization. However, the migration of services to the Cloud increases the vulnerability to existing IT security threats and creates new ones that are intrinsic to the Cloud Computing architecture, thus the need for a thorough assessment of Cloud security risks during the process of service selection and deployment. Recently, the impact of effective management of service security satisfaction has been taken with greater seriousness by the Cloud Service Providers (CSP) and stakeholders. Nevertheless, the successful integration of the security element into the Cloud resource management operations does not only require methodical research, but also necessitates the meticulous modeling of the Cloud security requirements. To this end, we address throughout this thesis the challenges to security evaluation and integration in independent and interconnected Cloud Computing environments. We are interested in providing the Cloud consumers with a set of methods that allow them to optimize the security of their services and the CSPs with a set of strategies that enable them to provide security-aware Cloud-based service hosting. The originality of this thesis lies within two aspects: 1) the innovative description of the Cloud applications' security requirements, which paved the way for an effective quantification and evaluation of the security of Cloud infrastructures; and 2) the design of rigorous mathematical models that integrate the security factor into the traditional problems of application deployment, resource provisioning, and workload management within current Cloud Computing infrastructures. The work in this thesis is carried out in three phases.