attributes is a critical concern. Various malicious entities can also monitor users’ behaviour and their access patterns to derive sensitive information about users and their attributes, which compromises their privacy.
Preserving user privacy, including data and sensitive attributes, is a significant concern while deploying Multi-Cloud. Therefore, it is vital to discuss the privacy term, its types, requirements, and metrics. Moreover, the different approaches used for preserving user privacy must be addressed from the state of the art, including their drawbacks and the potential for future improvement for more reliable and resilient techniques. Recent work, along with the implementation and research project, demonstrates that no particular approach or standard can address the privacy and security concerns of all cases. Rather, it depends on the context and privacy requirements in a specific application domain.
The major two areas that deploy Multi-Cloud are Genomics and Web domains. Genomic benefit from the federated cloud and web realizes the use of the cross-federated cloud to deploy the Single Sign-On (SSO). With regard to Genomics, it is critical to address different attack target user attributes and sensitive data—the necessity for the privacy-preserving technique that best fits the sensitivity nature of DNA. The open directions will be stated in this domain. For the SSO, there are different protocols that are deployed, including SAML and OpenID connect. The authentication and access tokens include sensitive information that could pose different attacks which threaten user identities and attributes. Moreover, most SSO protocols have some implementation flows and vulnerabilities. Thus, it is important to demonstrate the current practice of handling privacy in SSO protocols and highlight future directions.
In conclusion, the success of Multi-Cloud adoption toward building trustworthy environments is primarily driven by cloud user privacy and data security. The Multi-Clouds have no generalization for specific security and privacy-preserving approaches. Such techniques are primarily based on a particular context and the entities involved under a specific Multi-Cloud type and application. Enhancing the current cryptographic privacy-preserving techniques to adapt to the new level of Multi-Cloud evolution is necessary, as reducing its complexity and trade-offs between privacy, cost, efficiency, and scalability is a must.
Keywords:
Privacy-preserving, Multi-Cloud, Identity federation, federated cloud, cross-federated cloud, Single sign-on (SSO), Genomic data.
PhD Doctoral Committee:
External Reader: Dr. Mitra Mirhassani
Internal Reader: Dr. Saeed Samet
Internal Reader: Dr. Pooya Moradian Zadeh
Advisor(s): Dr. Ziad Kobti
