基于多关键字的Top-k布尔可搜索加密方案

针对在云存储服务中可搜索加密技术的研究，解决了现存3个问题：1)传统可搜索加密方法大多数只支持单关键字搜索，并且在安全索引过大时，搜索时间代价过大；2)现有方案大多使用倒排索引进行快速搜索，然而倒排索引并不支持动态关键字更新；3)现有方案大多数无法针对某些按需用户，根据关键字的重要性对搜索文件进行安全高效的top-k排序.提出了一种基于多关键字的top-k布尔可搜索加密方案(top-k boolean searchable encryption scheme based on multiple keywords, TBSE).该方案利用Goldwasser-Micalli与2DNF这2种加密算法构建了一种支持动态更新的安全索引；利用集合论与布尔搜索的相关知识构建了关键字交集索引与交集搜索令牌，以实现对多关键字的布尔搜索；利用TF-IDF加权技术与安全协处理器构造分数索引，进而能够对文件进行top-k排序.安全性分析表明：该方案保证了在已知密文模型与已知背景模型下的安全性.通过实验证明：该方案提升了对多关键字布尔搜索的效率以及索引存储的效率.

Top-k Boolean Searchable Encryption Scheme Based on Multiple Keywords

In view of the searchable encryption technology in cloud storage services, this paper solves the following three problems：Firstly, most of the traditional searchable encryption methods only support single keyword search, and when the security index is too large, the search time cost is too high; Secondly, most of the existing schemes use inverted index for quick search, but inverted index does not support dynamic keyword update; Thirdly, most of the existing schemes can’t sort the search files efficiently according to the importance of keywords for some on-demand users. In this paper, a top-k boolean searchable encryption scheme based on multiple keywords (TBSE) is proposed. In the scheme, Goldwasser-Micalli and 2DNF encryption algorithms are used to construct a secure, index supporting dynamic update. Based on the set theory and boolean search, keyword intersection index and intersection search token are constructed to realize boolean search for multiple keywords. The top-k sorting is realized by fractional index constructed by using TF-IDF weighting technology and security coprocessor. The security analysis shows that the scheme can fully guarantee the security of the given ciphertext model and the known background model. The experiments prove that the scheme improves the efficiency of multi keywords Boolean search and index storage.