Searchable Symmetric Encryption System Based On Graphic Database

Since the rise of cloud computing technology,cloud storage privacy issues to attract the interest of many researchers at home and abroad.To the cloud server does not trust as the prerequisite,encrypted data become mandatory.However,how to make encrypted data still be effective cloud computing has become a major problem,the most commonly used processing is keyword search.There are different solutions around this problem,but they are computational costs.Searchable encryption algorithm because of the efficiency and safety of comprehensive consideration,has been extensively studied.However,with the increase of users and the increase of the number of user files,the traditional algorithm which loads the whole index into the memory is a great challenge,and the efficiency of serial retrieval will be reduced.At the same time,the original encryption scheme does not consider the sustainability of the index.Aiming at the above problems,this paper proposes a searchable encryption scheme based on graph database,which transforms the memory index into a distributed graph node.As a new kind of distributed No SQL database,the graph database has excellent ability of graphics data processing and distributed storage,and has stronger ability of traversing graphs than traditional relational database.This has been the attention of researchers.This paper compares and analyzes the typical searchable encryption algorithm,summarizes the advantages and disadvantages of these works,and investigates the security level and formal proof method of the searchable encryption model and compared the performance of the graphic database and the traditional database in processing the graphic data by experiments.Based on the key words red and black tree,this paper proposes a searchable encryption scheme based on graph database,describes the process of index construction,parallel search and dynamic update,and gives the security proof.Finally,the API framework and system implementation are designed.By comparing the experimental system and the traditional model system,the model achieves the index persistence while the retrieval and update efficiency are superior to the traditional model.The validity and reliability of the model are dealt with in the large data performance test for the model.