Dynamic Access Control and Fine-Grained Searchable Encryption for Cloud Data Using Trust Evaluation and B-Tree Indexing

Abstract

The security management of data stored on cloud servers is of great significance, as it not only prevents data leakage but also ensures data integrity. Thus, to achieve secure management of data stored on cloud servers, the research starts with existing searchable encryption technologies and attribute-based searchable encryption technologies, and improved designs are made for both to form a complete encryption scheme. Firstly, in terms of searchable encryption technology, the study considers the dynamic changes in user permissions and constructs an encryption scheme that includes a trust value evaluation model and a dual dynamic access control mechanism. Among them, the trust value evaluation model needs to consider historical trust values, recommended trust values, and attribute related trust values, and then generate a comprehensive trust value to evaluate user permissions. Secondly, in terms of attribute-based searchable encryption technology, the research adopts a multi-branch balanced tree and linear secret sharing scheme to construct a fine-grained access control scheme. Among them, the B-Tree index structure is used to optimize search efficiency, and the linear secret sharing scheme is used to achieve access control for users. In the experimental part, the study uses Python programming language combined with PyCrypto and OpenSSL cryptographic libraries for testing to ensure the feasibility and performance of the solution, and the operating system is Windows 10. The results show that the maximum trust value calculation time considering the user permission scheme is 32 ms, which is 46 ms, 33 ms, 22 ms, and 19 ms lower than the maximum values of the four comparison schemes, respectively. In addition, the maximum access control determination time, CPU utilization, and memory occupancy of this scheme are 56 ms, 12.17%, and 13.95%, respectively. The maximum key generation time and communication volume for supporting fine-grained access control schemes are 180 ms and 3.257 Byte, respectively, and the average storage overhead for user keys and ciphertext is 3.94 KB and 5.37 KB, respectively. The encryption schemes designed by the research have good performance and can provide technical support for secure management of cloud server storage data without decryption for data queries.