Research On Multi-keyword Search Over Encrypted Cloud Data

In recent years,with the rapid development of computer network technology and storage technology,cloud computing technology has gradually become the mainstream technology widely concerned by academia and industry after distributed computing,parallel computing and grid computing.Cloud computing technology integrates and shares traditional IT resources,enabling users to use network resources on demand,which greatly reduces compute cost and storage costs.And because of the virtualization of cloud computing,each user can enjoy an independent high quality service.The convenience of cloud computing attracts more and more personal users or business users to join,which makes the cloud data explosively growing.Because most cloud service providers belong to private enterprises with commercial nature,they could be steal the privacy of users for some business purpose while providing services.Therefore,we not only focus on the storage and computing convenience brought by cloud computing,but also realize that the security threats for data privacy in public cloud storage environment.And it is becoming a problem that needs to be solved in cloud computing.Privacy data encryption is an important way to prevent privacy leakage,but this makes the keywords based search challenged.Therefore,the research of searchable encryption technology has become an urgent practical need.Based on searchable encryption technology,this dissertation research on the multi keyword search over encrypted data in public cloud storage environment,investigating the current research status and puts forward the following two solutions for the problems encountered.Searchable encryption which supports fuzzy search keywords has certain fault tolerance,it is more suitable for realistic search environment.Although there are some small spelling errors of the keyword entered by the user,scheme can still provide the correct search results for users quickly.Therefore it attracted many scholars to research.However,the existing fuzzy keyword ciphertext search scheme is not so efficient and does not support the dynamic update of the file.In view of above problems,this dissertation proposes an efficient and dynamic multi keyword fuzzy search scheme within encrypted cloud data,which can achieve fuzzy search of multiple keywords and support dynamic updating of files.The scheme utilizes locality sensitive hashing and bloom filter to generate index vectors and query vector.Based on these index vectors,a balanced binary tree is constructed as the index of the entire file set.In addition,by introducing the greedy depth first search algorithm,the related files are quickly searched on the constructed index tree.The search efficiency is further improved than the original fuzzy search scheme.The secure inner product method in computing k-NN problem is used to encrypt the index vector and query vector.Two schemes are designed to adapt to two different levels of security models respectively,and the security of the scheme is analyzed and proved respectively.Finally,comprehensive comparative analysis with other related schemes are carried out,and enough experiments are carried out.The results show that the search efficiency of the scheme is significantly higher than that of previous schemes.Search efficiency,accuracy,and security are important indicators of keyword search scheme.The existing multi keyword ciphertext search schemes adopt various index structures,such as inverted index,tree based index,to improve search efficiency,and enhance security by encryption algorithm and blind storage and so on.However,as for improving the accuracy,the importance factors and Semantic association of the search keywords are not fully considered.In order to solve this problem,this dissertation proposes a two-stage index based central-keyword ranked search scheme over encrypted cloud data.The scheme uses inverted index and forward index to build two stage indexes.Keyword weight algorithm is used to compute keyword weights for multiple search keywords,and takes the maximum weight as the central keyword.First,select the file set in the inverted index to narrow the search range,and then use the query vector and the index vector to calculate the inner product in the forward index to get the target file set.In this way,we ensure that the search results contain central key words,which improves the accuracy.In addition,by filtering on the inverted index in first search,a large number of vector inner product calculation processes are reduced.Therefore improve the search efficiency.As for security,inverted index is encrypted by pseudorandom function and symmetric encryption algorithm.The forward index is encrypted by the secure inner product method in k-NN problem,and the file set is encrypted by symmetric encryption algorithm.The scheme ensures that the privacy of the user's data and query is not leaked while improving the efficiency and accuracy of the search.