Efficient personalized search over encrypted data for mobile edge-assisted cloud storage

Abstract

Cloud storage services allow a data owner to share her/his outsourced data with other users, and enable the users to search target data by keywords. To ensure the data confidentiality, data owner always encrypt data using traditional encryption schemes before outsourcing. Whereas, it makes efficiently searching impossible. Symmetric searchable encryption (SSE) is a cryptographic primitive that resolves this tension. However, most existing SSE schemes do not consider the individual characteristics of users during the search, such that they cannot support personalized search services over encrypted data. Meanwhile, security and efficiency issues in the cloud service model have also severely affected the user’s search experience, and the introduction of mobile edge servers can solve these problems to some extent. In this paper, we propose a personalized searchable encryption scheme (PSED) for mobile edge-assisted cloud storage. Our contribution consists of three aspects. First, we incorporate the user’s preference factors into the user’s query which enable users to get accurate personalized search results. Second, the computational overhead of the cloud server is reduced by calculating the relevance scores of the subqueries and subindexes on mobile edge servers. Third, by cutting the index and the query matrix, the encryption efficiency of the index and the query matrix is improved. Security analysis shows that PSED can guarantee the privacy of the data and the user. Experimental results demonstrate that the proposed schemes are highly efficient and accurate.