Research On The Scheme For Structural Index Building Of Encrypted XML Data

XML has emerged as a new standard for data representation and data exchange on the Internet and more data is expressed in XML.Security is of paramount concern in an outsource environment,where a client may outsource data to a service provider that provides data management services as the service provider itself may be untrusted.Encrypting XML document offers a natural effective solution to preserve the confidentiality of the client's data.XML encryption not only can encrypt the entire document,but also can encrypt a certain element of the document and realize the fine-grained encryption finally.Now the challenge is to execute queries over the encrypted data quickly and accurately,without decrypting them at the server side.Building index for encrypted XML documents is an important method to improve query efficiency.In this paper,we design structural index of encrypted XML data based on B+ Tree for the purpose of efficiently querying the contents of XML documents,including structuring coding of XML elements,building B + tree index,designing B + tree index storage structure and query process processing.By performing encryption and decryption only on the client and not on the server to guarantee that neither document structure nor document content are disclosed on the server.In query processing,all the irrelevant elements are eliminated to reduce cost.According the result of experiment on XMark and DBLP datasets,our proposed scheme can be used to build B + tree indexes on different sizes and different types of data sets as wellAccording the result of experiment on XMark and DBLP datasets,our proposed scheme can be used to build B + tree indexes on different sizes and different types of data sets as well as can effectively and securely support queries against encrypted data sets.The advantage of building index on DBLP datasets is more obvious.Because the data structure of the DBLP tags are very few and the keywords to build B+ tree are relatively low,the efficiency of the query will be faster.