The Research And Implement Of Ranked Keyword Search Over Encrypted Data

To obtain both confidentiality and convenient utilization is always the goal of the searchable encryption schemes, which must be achieved and optimized gradually. Related works on searchable encryption focus on single-owner keyword search and rarely sort the search results. Moreover, the existing ranked Multi-keyword search schemes over encrypted data use much simple ranking functions and can hardly meet the effective data retrieval need.How to let the semi-trusted server store the owners’sensitive data safely and effectively search over them as well as rank the search results for the authorized users is the problem we need to solve.For the first time, we propose a secure inverted index(SII) which contains the relevancy scores between the search request and matching documents and implemet an effective and secure ranked searchable encryption scheme(SRSES) with the help of SII. SRSES allows more than one data owners to upload their encrypted files to the data server and search on them for authorized users later. During the process of search, the authorized user firstly provide the trapdoor of the request to the data server, the data server then execute the search operation on SII with the help of the received trapdoor. After that, the server rank the matching result securely and finally send the top-k encrypted documents that are most relevant to the search query to the user.In order to achieve multi-owner, multi-searcher and ranked search results, SRSES uniquely combine the following two advanced cryptographic techiniques:Homomorphic Encryption and KP-ABE. Homomorphic encryption enables the data server to compute on encrypted relevancy parameters and then rank the results without knowing the plaintext of the final relevancy score. With KP-ABE, SRSES can easily achieve secure, scalable and fine-grained data access control.