Project 1:

Systematic Evaluation of Cloud Computing Security

This research focused on developing measurable security metrics that will be used by cloud service providers, brokers, and consumers to assess cloud security services. The project presented several comprehensive security evaluation methodologies that enable service providers to guarantee security and transparency to users by deploying the necessary tools to continuously monitor the quality of their security solutions and predict future violations. The security evaluation metrics and techniques were used to develop a Security Service Level Agreement (Security-SLA) tailored for cloud computing based on a standard security vocabulary. The new SLA will speed up the cloud adoption process by facilitating the systematic comparison of cloud security offerings.

Project 2:

Towards Secure Interconnected Cloud Infrastructures

This project focused on security integration into the process of cloud resource federation to enable secure workload mobility in interconnected cloud infrastructures. A new cloud federation formation model was created by incorporating the Security-SLA to evaluate the security levels of service providers and their federations. The research leveraged powerful techniques from the field of cooperative game theory to enable service providers to collaborate in a security risk-aware fashion and even share their expertise in anomaly detection across interconnected systems to predict unknown attacks. Moreover, several solutions were proposed to securely manage the allocation of services and resources in multi-cloud settings.

Project 3:

Secure and Reliable Resource Allocation in Cloud Computing

This project was the first to integrate security requirements into the cloud resource provisioning and allocation processes, as well as mobile cloud computing. It focused on the design of security-aware resource allocation schemes that improve the security and resilience of cloud infrastructures. The research exploited the power of metaheuristics optimization and game theory to deliver centralized and decentralized solutions capable of securing large-scale cloud systems. It also contributed to the design of reliable cloud resource management schemes based on a quantitative, experimental evaluation of the reliability of cloud servers.

Project 4:

Robust Anomaly Detection in Cloud Computing

This research aims to leverage the advancement in the field of deep learning to design robust intrusion detection systems in cloud infrastructures and BigData analytics platforms. The research has already delivered promising outcomes and has now ramified in several new directions. For instance, my team and I are currently assessing the risks of black-box attacks on learning-based anomaly detection systems and exploring new ways to increase their resilience. Also, we are investigating the potential of integrating Federated Learning algorithms into existing intrusion detection solutions for collaborative and privacy-preserving security in edge computing.