30 / 2023-05-15 02:03:13

RESEARCH ON KEY TECHNOLOGIES OF DISTRIBUTED STORAGE OF COAL INTERNET BASED ON BLOCKCHAIN

Blockchain; Practical Byzantine Fault Tolerance(PBFT); Threshold Signatures; RabbitMQ; Distributed Storage

With the rise of digital currencies represented by Bitcoin. As the underlying technical support of Bitcoin, Blockchain technology has attracted extensive attention and discussion in the subsequent research process. Blockchain is widely used in finance, traceability and distributed storage due to its characteristics of decentralization and tamper-proof. As an important part of blockchain distributed storage system, consensus algorithm plays a role in maintaining data consistency and ensuring system security, and has a great impact on the performance of blockchain system such as transaction throughput and transaction confirmation time. Compared with public chain, alliance chain has fewer nodes and access mechanism, so PBFT consensus algorithm based on voting is generally used. However, there are still many shortcomings in the practical application of this algorithm. Firstly, in the three-stage consensus process, the communication between two nodes leads to high communication overhead and low algorithm consensus efficiency and poor scalability. Secondly, the primary node is selected in sequence, which has security risks. Malicious nodes may be selected as the primary node for several times, resulting in frequent view switching, which wastes system resources, and decreases system stability and reliability. In addition, the client only sends requests to the primary node. If a large number of requests are sent to the primary node, the processing performance of the primary node deteriorates, resulting in reduced system throughput and prolonged response time. Finally, there is no dynamic joining and exiting mechanism of nodes in this algorithm. The joining and exiting mechanism of nodes generally needs to be implemented by modifying the number of nodes in the configuration file, which will lead to the restart of the entire network and incur high overhead. To solve the problems of the above algorithms, an improved Byzantine consensus algorithm (TRPBFT) based on threshold signature and RabbitMQ is proposed. Firstly, the idea of grouping was introduced, and all nodes were divided into different groups according to their response speed to the group leader node, and the pairwise consensus between all nodes was changed into the pairwise consensus of the group leader node between each group. On this basis, the PBFT three-stage consensus process is optimized, and the messaging middleware RabbitMQ is introduced. In the pre-preparation stage, the client is decoupled from the primary node in the system, and messages are not sent directly to the primary node, but directly to the RabbitMQ node. The leader of each group is consumed by RabbitMQ subscriptions. RabbitMQ nodes will collect voting information and confirmation messages from other nodes, perform threshold signatures after verification, and then distribute the signed results to all nodes for separate verification. In this way, the traffic in the consensus process will be reduced. Improve system scalability and consensus efficiency as well as system response speed. Secondly, a reputation strategy is introduced to divide nodes into three types according to their reputation values: leader node, candidate node and ordinary node. Each group selects the node with high credit value as the leader node and uses the standby leader node mechanism to greatly reduce the overhead and security risks caused by frequent view switching caused by malicious nodes being selected as the leader node for many times, thus improving the system availability. Finally, experiments are designed for simulation analysis and verification of the algorithm. The experimental results show that the improved consensus algorithm has significantly improved the data throughput and consensus efficiency.