When data is outsourced at a cloud storage provider, data owners lose control over the integrity of their data and must trust the storage provider unconditionally. Coupled with numerous data loss incidents, this prevents organizations from assessing the risk posed by outsourcing data to untrusted clouds, making cloud storage unsuitable for applications that require long-term security and reliability guarantees. This project establishes a practical remote data checking (RDC) framework as a mechanism to provide long-term integrity and reliability for remotely stored data. At the same time, the project seeks to develop new functionality for remote data checking that overcomes limitations of early RDC protocols and improves the usability and deployability of RDC on existing cloud storage infrastructures. Unlike previous work, this research takes a holistic approach and considers RDC protocols that minimize the combined security costs of all data management phases over the lifetime of a distributed storage system. This includes prevention, repair, and retrieval. Maintaining the health of the data in a distributed storage system requires various transformations to be applied on the data and requires data to migrate among storage servers. This project develops novel RDC protocols that are compatible with the full range of replication, erasure coding and network coding operations employed by distributed storage systems, thus enabling owners to maintain better control over their data. This project increases the transparency of cloud storage platforms and improves the security dimension of storage outsourcing enabling wider adoption of cloud storage technologies. To disseminate these ideas, the project's educational activities include curriculum development, mentoring undergraduate and graduate students and engaging them into research, and outreach to high-school teachers.
