DMSE: Dynamic Multi-keyword Search Encryption based on inverted index

Abstract With the popularity of cloud storage, increasing people are willing to upload their files to cloud services. They will encrypt these files before uploading to protect the privacy of files. However, encryption makes effective search very difficult and inefficient. In this paper, we design a dynamic multi-keyword searchable encryption scheme on the encrypted cloud files, called DMSE. Firstly, the outsourced files in DMSE are pre-classified and preprocessed according to their privacy to achieve classified search. Secondly, based on the inverted index, DMSE realizes multiple keywords search, which protects the privacy of outsourced files and users and greatly improves the search efficiency. In addition, our DMSE extends the inverted index to the multi-DO/multi-DU scenario, which is the distinguishable difference between DMSE and other existing multi-keyword search schemes. Thirdly, our DMSE can further support the update of keywords to cope with the mismatch dilemma between the first extracted keywords and outsourced files, thus improving the practicality of DMSE scheme. Finally, comprehensive performance evaluation shows our DMSE scheme is effective and feasible in practical application.