Proofs of ownership on encrypted cloud data via Intel SGX
Department of Computer Science
To deal with surging volume of outsourced data, cloud storage providers (CSPs) today prefer to use deduplication, in which if multiple copies of a file across cloud users are found, only one unique copy will be stored. A broadly used deduplication technique is client-side deduplication, in which the client will first check with the cloud server whether a file has been stored or not by sending a short checksum and, if the file was stored, the client will not upload the file again, and the cloud server simply adds the client to the owner list of the file. This can significantly save both storage and bandwidth, but introduces a new attack vector that, if a malicious client obtains a checksum of a victim file, it can simply claim ownership of the file. Proofs of ownership (PoWs) were thus investigated to allow the cloud server to check whether a client really possesses the file. Traditional PoWs rely on an assumption that the cloud server is fully trusted and has access to the original file content. In practice, however, the cloud server is not fully trusted and, data owners may store their encrypted data in the cloud, hindering execution of the traditional PoWs. In this work, we make it possible to execute PoWs over encrypted cloud data by leveraging Intel SGX, a security feature which has been broadly equipped in processors of today’s cloud servers. By using Intel SGX, we can create a trusted execution environment in a cloud server, and the critical component of the PoW verification process will be executed in this secure environment (with confidentiality and integrity assurance). Security analysis and experimental evaluation show that our design can allow PoWs over encrypted data with modest additional overhead.
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Proofs of ownership on encrypted cloud data via Intel SGX.
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics),
12418 LNCS, 400-416.
Retrieved from: https://digitalcommons.mtu.edu/michigantech-p/14478