Multi-keyword ranked search with access control for multiple data owners in the cloud

Secure search is important for promoting the widespread use of cloud computing. As a common situation, when data users search for certain data from multiple data owners, searchable encryption raises the possibility of secure searches over encrypted cloud data. However, most of the existing schemes for the multi-owner model are based on asymmetric searchable encryption, which is inefficient for searching unstructured data. In addition, a Trusted Third-Party (TTP) is often required to assist in encrypting indexes or enforcing access control, which increases the cost and becomes a bottleneck to system security. To overcome these limitations, this paper proposes a secure multi-keyword ranked search scheme with access control for the multi-owner model, named MOMRS-AC. Specifically, we extend the basic secure KNN-based search scheme considering the single-owner model to an efficient multi-keyword ranked search scheme without TTP for the multi-owner model, and provide a signature algorithm to verify the search results. Also, we construct a fine-grained access control that supports recipient anonymity, large-universe and decentralized authorities. To lower the overhead of data users, we extend MOMRS-AC to support outsourcing of verification and decryption. Security and performance analyses demonstrate that MOMRS-AC is secure and efficient for the multi-owner model in practice.