| In cloud computing, to protect privacy when outsourcing database management systems, confidential data (e.g., bank account balance) has to be encrypted which however renders traditional query processing methods (e.g., the B-tree index) inapplicable. In this thesis, we consider processing range query search (e.g., retrieving all records with a bank account balance being less than some threshold) over a confidential numeric attribute with two main challenges. First, the confidentiality requires hiding the value and the relative order of the attribute in records from the cloud server, but computing a range query requires comparing the values of this attribute. Second, since the encrypted data does not preserve the closeness of plaintext values (for privacy reasons), a straightforward search has to scan the entire encrypted data to retrieve wanted records, which is not scalable for data of the cloud scale. In this work, we present a novel searchable encryption/index scheme to address these challenges. |
