| Blockchain effectively solves the trust problem between distributed nodes in the decentralized scenario,which has attracted extensive attention.With the continuous changes of blockchain application scenarios,the importance of access control has gradually emerged.Through the analysis of the access control schemes in the current mainstream blockchain platforms,it is found that the blockchain access control schemes also have some problems,such as coarse access control granularity,poor flexibility and difficulty dealing with the evil of nodes.To solve the above problems,this thesis proposes a modifiable blockchain access control scheme based on dynamic trust evaluation algorithm,which realizes fine-grained and dynamic access control on the blockchain,effectively prevents the malicious behavior of malicious nodes,and has higher security and better performance.The main research work of this thesis is as follows:(1)An algorithm is designed to dynamically evaluate the trust behavior of nodes and assign the trust behavior of nodes.The algorithm involves the calculation of equipment trust,transaction trust,incentive trust,indirect trust and comprehensive trust.In order to improve the accuracy of the algorithm,the weight of each trust degree is dynamically adjusted according to the node behavior.When the node has suspicious behaviors such as device attribute change,transaction failure and feedback failure,the algorithm will increase the weight of device trust,transaction trust and indirect trust in the comprehensive trust,and improve the impact of suspicious behaviors on the final trust,so as to detect suspicious behaviors more accurately and effectively solve the problem that it is difficult to detect malicious behaviors of nodes.Trust evaluation is carried out on the endorsement node selected for each transaction.By combining the evaluation results of each endorsement node,it can judge whether the node has access rights,and realize the decentralization of the execution of trust algorithm.In order to improve the data security of trust,the trust data is stored on the blockchain to realize the open,transparent and tamper proof storage of trust data,so as to ensure the security and consistency of trust data.(2)The chameleon hash function based on policy is used to replace the hash function on the blockchain to realize the correction of data transaction level on the blockchain.Users only need to replace the original ciphertext on the blockchain with a new ciphertext to modify the ciphertext access policy,which improves the flexibility of access control and better meets the needs of users.At the same time,the scheme will not produce new blocks when modifying transactions,and has high ciphertext storage efficiency.(3)Based on dynamic trust evaluation algorithm and modifiable blockchain,this thesis proposes a new access control scheme.The scheme adopts a double chain storage structure.The storage chain based on alliance chain is used to store ciphertext data,and the trust chain based on public chain is used to store trust data.Each time a node initiates a transaction,the blockchain network evaluates its trust and dynamically allocates access rights according to the node’s credibility,making the access control of the blockchain more flexible.The trust evaluation algorithm is used to measure the node behavior.When the node has suspicious behavior,it reduces its trust and limits the corresponding access rights,so as to avoid the harm of malicious nodes to the system.(4)Simulation experiments on the dynamic trust evaluation algorithm show that the algorithm can effectively limit the access rights of malicious nodes,timely find disguised malicious nodes,and improve the accuracy of malicious behavior detection.Finally,comparing this scheme with the relevant literature,it is found that this scheme has better security,higher access accuracy and lower ciphertext storage overhead. |