Use this URL to cite or link to this record in EThOS:
Title: Collaborative intrusion detection in federated cloud environments using Dempster-Shafer theory of evidence
Author: MacDermott, A. M.
ISNI:       0000 0004 6425 0148
Awarding Body: Liverpool John Moores University
Current Institution: Liverpool John Moores University
Date of Award: 2017
Availability of Full Text:
Access from EThOS:
Access from Institution:
Moving services to the Cloud environment is a trend that has been increasing in recent years, with a constant increase in sophistication and complexity of such services. Today, even critical infrastructure operators are considering moving their services and data to the Cloud. As Cloud computing grows in popularity, new models are deployed to further the associated benefits. Federated Clouds are one such concept, which are an alternative for companies reluctant to move their data out of house to a Cloud Service Providers (CSP) due to security and confidentiality concerns. Lack of collaboration among different components within a Cloud federation, or among CSPs, for detection or prevention of attacks is an issue. For protecting these services and data, as Cloud environments and Cloud federations are large scale, it is essential that any potential solution should scale alongside the environment adapt to the underlying infrastructure without any issues or performance implications. This thesis presents a novel architecture for collaborative intrusion detection specifically for CSPs within a Cloud federation. Our approach offers a proactive model for Cloud intrusion detection based on the distribution of responsibilities, whereby the responsibility for managing the elements of the Cloud is distributed among several monitoring nodes and brokering, utilising our Service-based collaborative intrusion detection – “Security as a Service” methodology. For collaborative intrusion detection, the Dempster-Shafer (D-S) theory of evidence is applied, executing as a fusion node with the role of collecting and fusing the information provided by the monitoring entities, taking the final decision regarding a possible attack. This type of detection and prevention helps increase resilience to attacks in the Cloud. The main novel contribution of this project is that it provides the means by which DDoS attacks are detected within a Cloud federation, so as to enable an early propagated response to block the attack. This inter-domain cooperation will offer holistic security, and add to the defence in depth. However, while the utilisation of D-S seems promising, there is an issue regarding conflicting evidences which is addressed with an extended two stage D-S fusion process. The evidence from the research strongly suggests that fusion algorithms can play a key role in autonomous decision making schemes, however our experimentation highlights areas upon which improvements are needed before fully applying to federated environments.
Supervisor: Shi, Q. ; Kifayat, K. Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
Keywords: QA75 Electronic computers. Computer science