Description

Recent years have witnessed an over increasing reliance on extern al services for data storage and management, towards the emerging cloud scenario, characterized by a rich and diverse availability of providers offering storage and computational functionalities. Together with data management functionality, the research and industrial communities have been investigating different solutions to ensure confidentiality of data whose management is outsourced to the cloud. Complementing data confidentiality, more recent approaches have also considered protection of access and pattern confidentiality, which require to maintain confidential to the server storing the data the fact that a given access aims at a specific target or that two accesses aim at the same target. Among these approaches, the shuffle index organizes data in a hierarchical encrypted data structure and provides access and pattern confidentiality by obfuscating accesses and dynamically changing the allocation of data to physical blocks, so to break the correspondence between data and locations where they are stored. In the existing system, the shuffle index provides a key-based hierarchical organization of the data, supporting then an efficient and effective access execution. These solutions however protect only the confidentiality of the data at rest. In this project the system proposed an approach that protects both the confidentiality of data stored at external servers and the accesses to them. The system build on such an indexing structure and on the idea of dynamically changing, at every access, the physical location of data, and provide a new approach to access confidentiality based on a combination of data distribution and swapping. The proposed system maintaining sufficient protection guarantees even in the presence of collusions among two, or even all three, of the involved servers.