Research On Ciphertext Retrieval In Cloud Storage Environment

With the development of cloud storage technology,more and more users choose to store their personal data in the cloud,which effectively reduces the burden of local data management,but the data stored in plaintext will bring serious privacy threats,and cloud data security and user privacy protection have become a hot topic of information security research.Using cryptographic algorithms to encrypt data and store it in ciphertext is an effective way to protect privacy,but users will encounter the challenge of how to search on ciphertext.Searchable encryption,as a cryptographic primitive that supports users to perform keyword search on ciphertext,achieves ciphertext search on the basis of ensuring data confidentiality,and has important research significance and application value for data security and privacy protection in cloud storage environment.However,most existing searchable encryption schemes still suffer from poor security,heavy certificate management burden,and difficulty in resisting quantum attacks.According to the above problems,this thesis focuses on searchable encryption algorithms to study the construction of two types of searchable encryption schemes,bilinear mapping and lattice on lattice,and the main contents and progress achieved are as follows:(1)In response to the key escrow and certificate management problems of traditional searchable public key encryption schemes,this thesis proposes a security-enhanced searchable encryption scheme without certificate,and by introducing the data owner’s private key in the ciphertext generation algorithm for authentication,the scheme can effectively resist internal keyword guessing attacks and ensure that keyword information will not be leaked even if the key generation center and the cloud server collude.Considering the data security threats brought by the untrustworthiness of the cloud,the scheme introduces blockchain to realize the trusted storage of keyword ciphertext and data hash to ensure their tamperability and traceability,and assists users in document integrity verification based on the key information stored on the chain.The scheme is shown to satisfy ciphertext indistinguishability and trapdoor indistinguishability in the face of attacks from two types of adversaries in the certificateless cryptosystem.The related analysis and performance evaluation results show that the scheme introduces a moderate overhead and is forward-secure.(2)In response to the security threats posed by the rapid development of quantum computing and the problem that most lattice ciphertext retrieval schemes only support single keyword or specified user retrieval,this thesis proposes a lattice multi-user ciphertext retrieval scheme that supports connected keywords.The scheme uses the lattice-based authorization algorithm and the lattice based randomization algorithm to embed the access set into the private key,and the correct retrieval can be achieved only when the access structure of the user’s private key and the ciphertext match,which effectively realizes the fine-grained access control in the multi-user scenario.To further improve the accuracy and reliability of the retrieval results,the scheme extends the single-keyword search mode to a connected-keyword search mode and uses decentralized ciphertext storage to ensure the security of keyword ciphertexts.The scheme is proved to be secure under the learning with errors assumption,and the performance analysis results show that the scheme has high efficiency.