High Performance Searchable Encryption System Supporting Multi-keyword Sorting And Its Application

With the development of intelligent terminals and mobile Internet,many users tend to store data in the cloud.Privacy-preserving in cloud computing is attracting more and more attention.In order to preserve the privacy of data owner,the data should be encrypted before they are outsourced to the cloud server,but traditional encryption limits the searchability of text data.In recent years,a number of schemes for searchable encryption have been devised to address the problem with information retrieval of ciphertext.But some of the current schemes cannot support keyword weight,some use linear search,and some do not support permission control.Most schemes for fuzzy search only support English.In short,the existing searchable encryption schemes,inadequate in terms of functions and performance,are confined to theoretical research and found lacking in application.In order to solve the above problems,we implement a high-performance searchable encryption system supporting multiple keywords and sorting,which can preserve users' privacy in the cloud environment.The main work of this paper is as follows:(1)To solve the problem of low-performance in the existing searchable encryption schemes,we propose a“high-performance secure KNN” algorithm.In this scheme we modify the secret key structure without reducing the security,thus greatly improving the performance.And we also furnish a detailed proof of security.(2)We attempt to combine the scheme proposed by Xia et al.with the proposed “high-performance secure KNN” algorithm to effectively improve the speed of cipher-index construction,and we use vector space model and TF-IDF value to build tree index,which reduces the search complexity to the sublinear level.(3)In order to remedy the deficiencies of cipher-index search algorithm in the worst case,which entail a drastic reduction in search performance and accuracy,we optimize “Greedy Depth First Search(GDFS)” algorithm and propose “Enhanced Greedy Depth First Search(EGDFS)”algorithm,which realizes accurate and efficient search in any situation.(4)In order to achieve sorting,we sort files in the result set by similarity.(5)In order to solve the problem of user error input by five-stroke input method,we implement Chinese fuzzy search by shape-nearwords table on the ciphertext domain.In addition,we carry out a series of experiments to illustrate the efficiency of our scheme and contrastive scheme at all stages.Performance comparison results show that our scheme can effectively enhance the performance of key generation,cipher-index generation,trapdoor generation and search.Integration test result shows that the proposed system can protect users' privacy in the cloud environment.In the end,we illustrate the proposed system may be used in the field of secure cloud storage and group sensing to solve the problem of privacy-preserving.