Research On Searchable Encryption Over Privacy Information In Medical Cloud

Nowadays cloud computing and big data are the hot research fields of information technology.Cloud storage,an important component of cloud computing,is chosen by more and more users,due to its enormous advantages,such as convenience and low cost.For the medical cloud system,as the cloud service provider is the administrator of medical big data,there are great challenges for users' information security from the cloud administrators and the malicious attackers.In order to ensure the security and privacy of medical big data,users usually encrypt sensitive data,and then upload them to the cloud server.Classic encryption technology can protect the security and confidentiality of data,but searching over the encrypted medical big data becomes a problem.Therefore,it is very urgent to study the storage structure,data characteristics and security requirements of the medical cloud,and further propose secure and efficient searchable encryption schemes for medical big data.In order to search over the encrypted medical big data in medical cloud system,the research status of searchable encryption technology in medical cloud storage environment is analyzed comprehensively and systematically.Three search methods in the encrypted data are studied including the keyword fuzzy search,multi-keywords fuzzy search,and search supporting conjunctive queries.Several secure and efficient searchable encryption schemes for the medical big data in different medical cloud models are further designed.Three research contents in this work are as follows.1.An efficient single-keyword fuzzy searchable encryption scheme in hybrid medical cloud is proposed.In order to meet the requirements of single-keyword fuzzy search in the hybrid medical cloud,which is widely used by medical organizations,a ciphertext keyword similarity measurement is proposed.The editing distance and Jaccard similarity coefficient are used to establish the “similar coordinates” of keywords,so that the keyword's fuzzy matching problem is converted into a similar distance calculation to determine the similarity between the two keywords.the secure indexes for “similar coordinates” are built by using an improved mutable order-persevering encryption(MOPE)based on KD tree.The nearest value query algorithm is used to achieve fast fuzzy matching.In the proposed scheme,keyword fuzzy search is achieved by keyword similarity measurement,and the search results are returned with ranking according to the relevance between keywords and files.Both the search efficiency and communication cost are reduced.The analysis shows that it satisfies the adopted adaptive semantic security.2.An efficient multi-keyword fuzzy searchable encryption scheme in public medical cloud is proposed.In fact,there are too many results returned for search over medical big data in public medical cloud,which leads to a decrease in the accuracy and efficiency of search.Keywords are mapped to spatial vectors based on the spatial transformation mechanism,and then secure indexes are built over the transformed vectors using Locality-Sensitive Hashing(LSH).These make a new multi-keyword similarity measurement.In our scheme,Bloom Filter is used as a keyword conversion mechanism,and the property of the Minhash function are used to achieve the fuzzy matching of the converted keywords.The returned results are sorted according to the accumulated value of the relevance between keyword and document.As the number of returned results is limited,the accuracy of search is improved.At the same time,the proposed scheme can use different keyword conversion mechanisms and corresponding LSH functions according to the application,which increases the flexibility of this scheme.Security analysis shows that the proposed scheme satisfies adaptive semantic security and reduces information leakage.3.A multi-user non-interactive searchable encryption scheme that protects query privacy in a complex medical cloud is proposed.In order to meet the requirements of multi-user non-interactive and efficient search in the complex medical cloud,searchable encryption supporting conjunctive queries can be used to reduce the search scope and improve search accuracy.Using trapdoor generation algorithm based on secure computing and attribute-based encryption,searchable encryption supporting conjunctive queries is extended from single-user to multi-user,which protects the privacy queries with no interaction between the user and the data owner,so it is unnecessary to be online for the data owner.The security analysis shows that this scheme can resist malicious attacks from the servers,authorized centers,and unauthorized users,and it satisfies adaptive semantic security with different adversaries.Finally,the main achievements of this work are summed up comprehensively,and the directions for further research are pointed out as well.