Ciphertext Fuzzy Search Method Based On Multi-Keyword

With the rapid development of data outsourcing technology,more and more users begin storing data to a third party server,in order to reduce the burden on the local management of data.However,the server is not authentic or semi-authentic,many store their data with ciphertext in order to ensure data security.However,conventional encryption techniques will result the data lose many original features,while ensuring the confidentiality of the data,their common functions will not be providing,such as data sharing and distribution,keyword search with ciphertext data,etc.As a result,How to effectively search ciphertext stored in the server has become an urgently issue.In this thesis,the service provider is considered as a credible entity,this thesis propose a multi-keyword fuzzy search method on the cyphertext,around ciphertext data search problem.The method cited bloom filter and location-sensitive hash function and so on.When users want to upload a file,they should firstly convert them to vector using dual encoding functioin,and then use the location-sensitive hash function mapped them to a Bloom filter.What's more,when users want to search a file,they also should transformation query-keyword on the same way.To server,computing the inner product of the safety index parameters and the trapdoor,then according to the result of calculation,return the target file to the user,is the process of search.In terms of security,through the interaction between the offensive game that simulates users and servers use random parameters,it proves the method is choose keyword attack security with the random parameters and the offensive game.Finally,this thesis use Crypto API and Openssl to design and implement the multi keyword fuzzy search on the ciphtext system.First,designing the architecture of the system,then elaborating the document processing,key processing and ciphertext search;then proposing the definition of the funcition of the system,and the method of implementing the prototype system,respectively testing its functionality and performance,and then analyze the results of the test.The results show that the proposed method is feasible.To the nature,setting the optimum parameters for the system by testing the performance impact of various parameters,such that minimized the error of the search results.