Privacy-Friendly Protocols for Cloud Security

The increasing use of untrusted ICT-based environments for storage and processing demands for effective access control and protection mechanisms for the outsourced user data. Failure to secure the user data results in eliminating the benefits of such services. Cloud computing has several advantages like cheap service costs, higher computing power, scalability and higher performance but there are security issues for cloud providers as well as customers.

With the advancement of genome sequencing technologies, the need to securely store and process the genomic data is high. DNA data leakage is a serious issue because once leaked it cannot be resolved as the genomic data of an individual does not change. It also has the risk of leaking information about his ethnic heritage and relatives (without their consent). Similarly, an individual's financial information is sensitive as it could leak information about his activities and other personal details. The enormous size of data in such applications has to be considered while developing a feasible privacy-preserving method to handle data in the cloud.

In this era of online privacy erosion, it is highly important that there is an end-to-end protection of data to preserve user's privacy and to ensure minimal data leakage. This PhD aims to develop protocols for protecting data in the cloud driven by the privacy-and-security-by-design principles.

First we will investigate the state-of-the-art techniques for handling private data and the level of security offered by them so that the shortcomings can be identified. Next, privacy-friendly solutions for the identified security requirements will be devised, taking into account all the actors involved in the process. This study will also focus on developing secure and efficient cryptographic building blocks and investigating the tools for data processing in the encrypted domain. We will utilize techniques like homomorphic encryption to prevent data from being compromised even when it is accessed and operated upon by multiple entities. We will implement the developed methods to demonstrate the efficiency and improved security of the system.