摘要:
Web search engines accumulate substantial quantities of personal data extracted from the queries of internet users. These datasets encompass a range of sensitive factors, including in- dividual interests, engagements, health issues, work chronicles, religious principles, and political orientations. This aggregation of data components heightens the threat to user privacy, especially in situations characterised by data breaches, unauthorised access, or inappropriate exploitation of their private information. The present thesis introduces a unique solution Privacy Extension for Search Engine (PESE), which combines the concepts of unlinkability and indistinguishability to enhance the privacy and anonymity of web users. Our solution employs a Trusted Execution Environment (TEE) to ensure the confidentiality of search queries, thereby thwarting any potential monitoring or interception by search engines and external entities. TEE consists of a secure processor component that offers selected programmes executing on the processor with data security, secrecy, as well as integrity. It effectively isolates these applications from the Rich Execution Environment (REE). We utilize Enarx, an open-source deployment framework for designing and building our TEE application. It provides a robust and flexible environment that facilitates the secure execution of workloads within a TEE. Enarx's flexibility in accommodating different hardware architectures and platforms allows our PESE solution to be easily deployed across diverse systems without signif- icant adjustments. This wide-ranging compatibility increases the applicability and accessibility of our solution, making it suitable for a wider array of environments. Throughout this thesis, we expand on the implementation of PESE via TEE, elaborating on its operation's complexities. Furthermore, we conduct a thorough review of the Enarx deployment framework, highlighting its central role in the design and development of our TEE applicati