| The security sharing of data is very important to give full play to the value of data elements.Because data is easy to copy and what you see is what you get,it is difficult for data owners to control the data after sharing.In actual scenarios,data is often generated by multiple parties and there may be multiple data owners(hereinafter referred to as "data owners").Different data owners have different requirements on data access control,which are manifested in the access control strategy and control granularity.However,the existing research has not discussed this issue sufficiently.Based on blockchain,attribute-based encryption and other technologies,this paper studies multi-owner data access control strategies and methods to ensure joint authorization of multi-owner data sharing,data finegrained access control and automatic decision and strict implementation of access control policies.The main research contents and contributions are as follows:1.To solve the problem that the access policies formulated by different owners may be inconsistent,this paper proposes an attribute based access control policy fusion method.Firstly,the dual owner strategy fusion algorithm is proposed for the data with dual owner.When the access strategies formulated by the two owners of the data conflict,the balance matrix is calculated using the mutual influence weight and willingness score of the two owners to resolve the conflict.On this basis,a more complex data multi-owner access control policy fusion problem is studied,and multi-owner joint authorization of shared data is realized without conflict.After the initial pair fusion of each owner strategy,considering the more complex interaction relationship among multiple owners,the influence score matrix composed of contribution decision degree,direct decision degree and indirect decision degree was reconstructed to ensure the accuracy and fairness of the balance matrix.2.Aiming at the problem of fine-grained access control of data sharing,a fine-grained access authorization scheme based on cloud blockchain is proposed.By reducing the scope of data access successively,direct ciphertext retrieval of the whole blockchain is avoided,ensuring secure access and effectively improving access efficiency.In this scheme,the shared data is encrypted and stored in the cloud,and the index mapping table of the ciphertext and the address on the blockchain is designed.Firstly,according to the request data type specified by the data requester,the index mapping table is used to conduct a search on the cloud server,and the scope of data access is narrowed initially.The information obtained in one retrieval is used to determine the access control policy based on attribute encryption,and the on-chain address of the key index meeting the requirements of the specific attribute access policy is obtained.Finally,the actual ciphertext storage address is obtained by keyword matching with the address on the blockchain.3.The cloud is generally responsible for the existing access control based on attribute encryption,and centralized access control policies may not be strictly and effectively implemented.To solve this problem,this paper proposes a multi-owner joint decision scheme of access control policy based on threshold password.In this scheme,the converged attributebased access control policies are written into PIP,PAP and PDP smart contracts and deployed on the consortium blockchain with multiple owners as nodes.By combining the threshold cryptography mechanism with Execute-Order-Validate consensus pattern,nodes decide whether to give access key shares to participate in policy association decision according to the results obtained from executing smart contracts of access policies.The final access control decision is generated according to the common wishes of most owners to ensure the joint decision and strict implementation of multi-owner integration access control policies. |