CloudDBGuard: Implementing cryptography-based approaches to secure data management in cloud databases

• Applicant: Professorin Dr. Lena Wiese
• Subject Area: Security and Dependability, Operating-, Communication- and Distributed Systems
• Term: from 2013 to 2023
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 244972644

Column family stores comprise a group of non-relational ("NoSQL") databases. Because of their flexible and efficient behaviour, column family stores are widely used by cloud storage providers, where several users can rent storage space. This raises concerns how confidential data can be protected from a curious cloud storage provider or other attacks of unknown third parties.So far, column family stores do not offer any means to ensure the protection of confidential data from unauthorized access. The main goal of the FamilyGuard project is hence to leverage secure and user-adaptable encrypted storage of data in column family stores. At the same time, as a side condition, the technical procedures of column family stores must be maintained and an efficient query processing by the database server must be supported. Due to this a sensible balance between two contrary requirements must be established: providing security guarantees as good as possible while at the same time managing data as efficiently as possible. This goal is be achieved by the FamilyGuard project by harnessing searchable and order-preserving encryption for column family stores. A proof-of-concept implementation will be done for column family stores that are freely available as open source software and are hence extensible in this research project; a performance analysis will be based on the Yahoo Cloud Serving Benchmark.

• DFG Programme: Research Grants