An Effective Verifiable Database Protocol in Severe Untrusted Environment

The security of outsourced data has been a key issue in cloud applications. Previous researches have focused on the remote data auditing and verification schemes. Although some of them have the properties of dynamic update and public audit, most of these schemes implicitly assume an honest client. Actually, the client has the motivations to corrupt the verification results against server, and the server may suspect the verification results under the situation where the client is dishonest. Up to now, few solutions have been proposed when both the client and server are not fully trusted and either side could be malicious. In this paper, a strict verifiable database (sVDB) protocol is proposed to tackle this problem. The concrete construction based on the bilinear pair technique can support client's real-time query and batch update operations. We prove the security of the scheme by contradiction under the square computational Diffie-Hellman assumption. We also implement the protocol and carry out several experiments to compare its performance with the state-of-the-art VDB scheme using vector commitment. All the results show that the proposed sVDB scheme achieves good security performance and high computation efficiency.