Research On Attribute-based Encrypted Efficient Multi-keyword Retrieval Mechanism In Multi-cloud Environment

With the rapid development of cloud computing technology,more and more companies are offering cloud services.Since a single cloud service may lead to a single point of failure due to problems such as network congestion or hacking,a multi-cloud computing model has started to be proposed and widely used.The multi-cloud model can provide more flexible and reliable services to users but also brings more complex security issues.How to retrieve files with keywords from a large amount of encrypted data in multiple cloud servers while maintaining high security has become a popular research direction in industry and academia.The existing searchable encryption schemes still need some improvements,such as inflexible access control policies,low search efficiency,support for only single keyword retrieval,and difficulty applying to multicloud environments.To address the above problems,this paper proposes an Attribute-Based MultiKeyword Retrieval Scheme in Multi-Cloud Environment.The scheme is based on an efficient attribute encryption algorithm and designs a two-segment retrieval trapdoor and a secure index containing encrypted keyword sets to realize multi-keyword search.Moreover,a retrieval server is introduced to search encrypted files from multiple cloud servers quickly and accurately based on its stored access control structure and index information.However,users need more fine-grained retrieval control in some scenarios while ensuring specific retrieval efficiency.Nonetheless,the lack of logical connections in the multi-keyword set used in the above scheme makes it difficult to satisfy users’ expressive search.This paper proposes a Hybrid Boolean Multi-keyword Retrieval Scheme Supporting Fine-Grained Retrieval Control in Multi-Cloud Environment to address this problem.Hybrid Boolean keywords are introduced in this scheme,i.e.,users can search using more expressive Boolean keywords,such as the access tree structure of keywords.In addition,each user holding a user token can delegate keys for another attribute set for a subset of its users.Finally,the security analysis verifies the security of the two schemes in this paper under known context and known ciphertext models with confidentiality,resistance to collusion,and fine-grained access control.The performance evaluation shows that both schemes have lower time and storage overhead in the retrieval trapdoor generation and retrieval matching,which are the most frequent operations,compared with existing schemes while ensuring high security.