BFDup: Batch Fuzzy Deduplication Scheme for Massive Data in Non-Trusted Environments

Existing fuzzy deduplication solutions focus more on providing secure and low redundancy storage services while ignoring the efficiency of deduplication matching-batch checking, which is not supported. In the traditional fuzzy deduplication, the system finalizes the deduplication check by matching deduplication labels stored in the server and uploading one from the client. It can be learned that such a client-based deduplication approach relies on a one-to-many matching mechanism, which is extremely inefficient when dealing with massive outsourced data on a large scale. In contrast with conventional matching, we propose a more efficient and secure server-based batch fuzzy deduplication mechanism in this paper. Our solution adopts many-to-many batch deduplication judgment to accomplish high-efficiency deduplication of encrypted data, as well as improving the security of the system by preventing collusion between servers.