Study Of Secure Ciphertext Retrieval Research In Cloud Environment

With the rapid development of network technology and applications, the data generated by the network grew sharply. Then the data storage technology and application of technology has also been rapid development, In particular, the development of cloud computing technology and data mining technology.Cloud computing technology makes it possible to store large amounts of data to the cloud, it improves the utilization of network equipment and efficient management. But it also brings security issues, the security of data becomes critical. Searchable encryption technology can let users to get ciphertext data from cloud fast and secure. Currently, Searchable encryption technology available based on encryption type are divided into SSE and SAE, for the structure of scheme, it is divided into a searchable encr yption base on liner search and index-based searchable encryption scheme. Big data can be used as data mining resources. We can find the potential knowledge from ciphertext retrieval historical data collection using data mining technology. For example, the degree of association between ciphertexts, the closeness between the trap-doors and the tightness between trap-doors and ciphertexts. Existing ciphertext retrieval scheme can not resist the new data mining attack.This paper studied deeply on searchable encryption scheme and data mining. Identify existing schemes have security issue, when they facing data mining arracks. And we make the appropriate security improvements. The main work in the following areas:Firstly, based on the study about data mining technology, we find data mining techniques shortcomings. The anti- mining attack algorithms security definition is given. On this basis, the characteristics of ciphertext retrieval are also combined. We define the ciphertext retrieval anti-mining security.Secondly, according to the definitions, We proposed an safer ciphertext retrieval scheme. There are three major improvements: 1.the search key is made random, as we put a random bit string in it. 2. Retrieving trapdoor is double encrypted by using random search key. So the tarp-door is safer than before. 3. Search server to perform pro xy re-encryption. Then we can use the trap-door to retrieve.At last, To achieve improvement scheme and to test it using the test data. First, we test and verify the statistical randomness of RSKey and trap-door T. Then, we compare the retrieval efficiency between original scheme and our improved scheme. According to the results, our improved scheme is safer and only a small decrease in the retrieval efficiency.