Centralized keyword search on encrypted data for cloud applications

Centralized approaches are widely adopted to improve the qualities of outsourced services owing to its feature of efficient system management. Because the cloud is untrusted, data are usually encrypted before outsourcing. One of the interesting applications is searchable encrypted keywords, a well-known cryptographic primitive that allows encryption while enabling search for keywords. However, achieving data retrieval without revealing privacy in a cloud-based centralized system is still a challenging problem. In this paper, we introduce centralized approaches to searchable encryption and present a novel centralized system for retrieval services. In our system, the centralized manager can search and access all the encrypted data from authorized users, while each user can only search and access his or her own data. The system builds on a new cryptographic notion called centralized keyword search on encrypted data. We formalize its security model and propose a centralized keyword search on encrypted data construction that is featured by short ciphertext and search result verification. We further extend the construction for removing secure channel and enabling batch authentication on data legalities. The experiment demonstrates the performance of our proposals.