Privacy-Preserving Ranked Multi-keyword Fuzzy Search on Cloud Encrypted Data Supporting Range Query

Abstract

It is a desirable technique for cloud users to make the fullest use of cloud encrypted data by searching what they need through input keywords. Exact keyword search schemes over encrypted data have been well tackled for better retrieval efficiency and accuracy. However, existing researches on fuzzy keyword search are mainly based on single-input keyword, where multi-keyword fuzzy search remains to be unsolved, and keyword-based search application expansion, i.e., range query based on fuzzy search, has not yet proposed. In this paper, for the first time, we propose a novel ranked multi-keyword fuzzy search scheme supporting range query called RMFSSRQ by exploiting order-preserving encryption and locality-sensitive hashing. Our scheme achieves ranked fuzzy keyword matching by algorithmic design to support retrieval ranking of returned encrypted files. It can also conduct fuzzy search without pre-defined keyword dictionary restraints and eliminate increasing computation and search overheads of multi-keyword fuzzy search compared with traditional fuzzy keyword search schemes. As an expansion of our scheme’s application, range query can be achieved by building secure per file Bloom Filter (BF) index. So, the scheme can achieve ranked multi-keyword fuzzy search as well as range query on cloud encrypted data through two-layered BFs per document. Extensive security analysis and experimental results on real-world data set show that our proposed scheme can securely reach the design goals for keyword search on encrypted data. To the best of our knowledge, this is the first try to achieve ranking of retrieval results and range query based on fuzzy search over cloud encrypted data.