Research On Searchable Encryption Technology Based On Cloud Storage

In recent years,the amount of data has been increasing linearly,and local storage has long been unable to meet the user's storage requirements for data.More and more users choose to store their data in the cloud depending on the rapid development of the cloud computing technology.A data security problem may be caused is that the data can be easily stolen by criminals and it brings losses of users' interest,for that the users loss control of their data.With more and more data stored in the cloud,the retrieval of some target data becomes a thorny problem.It can not satisfy users' retrieval needs by the way of downloading all of the documents and retrieving after decryption in each retrieval process.Therefore,the searchable encryption technology came into being.The searchable encryption technology can directly retrieve the target data in the ciphertext state,and after decryption,the target data in plaintext state are sent to the user,which saves a large amount of local storage and bandwidth computing resources.At present,some searchable encryption technologies are mostly focused on keywords,multi-users and other directions to improve retrieval efficiency and security,and there is no certain solution in the direction of multi-server.Besides,it can not be ignored about the relationship between words.Although some studies have introduced the concept of phrases,the research still has not become a hot research spot.In view of this,on the basis of previous studies,this thesis proposes two different searchable encryption schemes based on the existing application requirements.The first scheme proposes a multi-server ciphertext retrieval method based on DHT.The second level hash table is constructed by the universe has function as the data index structure,and the storage address of the file is stored at the mapping location of the hash table.The distributed storage of files on multiple servers is handled by a consistent hash algorithm.The second scheme proposes a multi-server searchable encryption scheme based on concept hierarchy.It designs the respective index of sub-servers and documents using the concept hierarchy tree,which can locate the most relevant cloud servers in order to improve the retrieval efficiency.And meanwhile it introduces TF*IDF method to further enhance the accuracy of retrieval.Simulation experiments for the two schemes and demo for the second scheme are implemented.The experimental results show that the proposed schemes improve the retrieval efficiency and the accuracy.