Secure and flexible data sharing via ciphertext retrieval for cloud computing

Abstract

Cloud-based data storage is attracting more and more users, yet there is no well-recognized solution that satisfies their requirements for both secure storage and flexible sharing. Many researchers have proposed different schemes, but none of existing schemes accounts for security of cloud data, flexibility of sharing, computing cost, and other factors at the same time. In this paper, we propose a secure and flexible data sharing scheme for cloud computing. Combining homomorphic encryption and a TF-IDF based vector space model, we apply ciphertext retrieval in cloud data sharing for the first time. Adapt to the distributed nature of cloud computing, we propose a parallel homomorphic encryption algorithm based on MapReduce. In addition, we also adopt CP-ABE in our scheme to offer fine-grained access control according to appropriate access structures. We have conducted several experiments on a Hadoop cloud platform to evaluate the proposed scheme. The results show that our data sharing scheme has the advantages of higher security, flexibility, and efficiency with less computation cost.