Research On Multi-keyword Searchable Encryption Technology For Cloud Storage

The development of cloud computing and cloud storage services has brought great economic benefits and research value to society.Data Storage-as-a-Service can greatly reduce the cost of data management by providing expertise and maintenance services to customers with limited resources.The storage method of storing data in the cloud server saves the user's local storage space,but the cloud server is not completely reliable for the user,which will cause privacy disclosure.To solve this problem,searchable encryption(SE)scheme has been proposed in recent years.The SE enables legitimate users to retrieve encrypted data according to certain rules in the cloud server and does not leak sensitive information about encrypted data.Fuzzy keyword searchable encryption schemes take into account the minor clerical errors of users.When the query keyword does not match the index keywords stored in the cloud server,the fuzzy keyword searchable encryption schemes can return the matching encrypted documents.In addition,multi-keyword searchable encryption is more meaningful in reality.Multi-keyword searchable encryption can generate a common index for multiple query keywords,instead of searching for each keyword separately and then finding the intersection.This paper studies multi-keyword fuzzy searchable encryption,the main work of this paper includes the following two aspects:(1)A new index construction method is proposed for multi-keyword fuzzy searchable encryption scheme.This method is based on vector space model to process keywords.In the process of constructing the index,tools such as locality sensitive hashing and Bloom filter are used.This method can identify more spelling errors,for example,write one letter more or less,write one letter wrong,and reverse the order of two letters.And the experimental results show that the search accuracy of this scheme has been improved.When generating trapdoors,random permutation functions,Euclidean norm and Jaccard similarity coefficient are used to achieve probabilistic trapdoors,thereby avoiding the leakage of search pattern.(2)In view of the low search efficiency of the multi-keyword searchable encryption scheme,this paper used a two layer index structure in the scheme,namely the inverted index and the forward index.The use of inverted index improves the search efficiency of the scheme,while the forward index enables the scheme to support multi-keyword fuzzy search.And a proxy server is introduced to prevent more information leakage.The experimental results show that this scheme greatly improves the search efficiency of the multi-keyword searchable encryption scheme.In addition,aiming at the problem that symmetric searchable encryption scheme does not support authentication,the scheme proposed in this paper uses certificate-less encryption,so that the symmetric searchable encryption scheme can realize user authentication,only users who meet the conditions can get the encrypted document and decrypt it.