Research On Confidentiality Protection Technology In Outsourced Database

Outsourced Database (ODB) is an application of Software as a Service (SaaS) delegating database owner’s (DO) tasks of creating, managing, updating and supporting query to outsourced database service provider (DSP). With the rapid development of Cloud Computing, which serves as application platform of ODB, surveyors pay more and more attention to ODB to achieve it a secure, reliable and efficient service. Because the DSPs in ODB scenario are not absolutely trusted, they are always honest, but also curious, so the owner’s database is deployed on an untrusted server, which differs greatly with conventional trusted database server. The traditional measures to ensure database confidentiality are not always available now. The confidentiality of ODB is confronting an increasingly challenge.Based on the analysis of existing ODB work, this paper proposes a wrap-around partition cryptography indexing technology for encrypted data and two outsourcing schemes-for relation database and XML database using attribute fragmentation and polynomial splitting respectively. The works that have been finished are as follows:1. Overview the current achievements and trends about confidentiality protection of ODB, outline the goals of confidentiality and the obstacle and technical haze in the researching work. Introduce a confidentiality protection architecture involving Trusted Third Parties (TTP) for ODB, which is the model foundation of following work in this paper.2. Propose a wrap-around petition cryptography indexing (WPCI) scheme with AES and modular arithmetic. With the concept of secure parameter, expend the partition range which the attribute value can map to, in which way can the confidentiality of index itself is enhanced. This method decreases the number of false positive tuples in the results set of DSP responded to equal queries, while, doesn’t cause a mass of false positive tuples increasing in the results set of DSP responded to range query. So, this scheme achieves efficiently supporting ability both to equal query and range query at the same time.3. In order to decrease the encryption and decryption overloads of DO or TTP, a relation database outsourcing scheme using attribute fragmentation and WPCI is proposed. Firstly, design an application oriented strategy to compute attribute fragmentation cost, which concerns about the query type and the position that the involved attribute in. Secondly, using the self-adaptive strategy, establish an algorithm to find the least cost attribute Fragmentation scheme for given table. Furthermore, break the link of identifiers of the same tuples in different fragmented attribute sub-set to resist DSPs-Collusion, and establish WPCI index on the attribute within encryption attribute set. Finally, give a brief look of the query executing mechanism for this scheme and demonstrate an experiment to test its efficiency.4. Propose a confidentiality protection scheme of Outsourced XML database using improved polynomial splitting. To avoid the extra overload of computing its children value in verifying node value in existing scheme, this scheme add additional assistant verify data structure to each non-leaf node of the XML elements tree expressed with polynomial, which achieves node value verifying without the presentation of node’s children value at a low cost of pre-processing. Furthermore, the query executing method is carried out in our improved verifying situation. The final experiment results show our contribution in speeding up the query executing.