Research On Searchable Encryption Technology And Its Application In Cloud Storage

With the development of cloud storage,more and more data have been migrated to the public cloud.It facilitates efficient storage management and obtaining more convenient file services.To ensure the security of sensitive data,they are encrypted before being uploaded to the cloud.However,the original logical relationships between documents are destroyed~which makes search on encrypted data difficult.Due to the importance of information retrieval.,searchable encryption has got extensive attentions in recent years.In order to further improve the efficiency and accuracy of searchable encryption,we mainly do the following research xworks:(1)The full-text search framework Lucene is introduced into the searchable encryption schemes,and an improved probability model BM25L is adopted to replace the relevance score model of the Lucene system.Experimental results show that this method can improve the precision and accuracy of the search results.(2)To improve search efficiency,an improved clustering algorithm CAK-means is utilized to build clustering index on documents.The optimized clustering algorithm can improve the relevance between documents,so as to achieve the goal of improving search efficiency and accuracy;(3)Since the searchable encryption schemes are closely related to the applications of reality,we try to apply the searchable encryption scheme to the personal health record(PHR)system.A searchable encryption scheme for PHRs based on the NoSQL database,MongoDB,is proposed.Experimental results demonstrate the feasibility of searchable encryption schemes in real applications.Finally,the related research works are summarized and the unsolved problems and future research direclions arc stated.