Research And Application Of Multi-signature And Flexible Byzantine Fault Tolerance Consensus Algorithm In Blockchain

Blockchain technology has revolutionized some traditional business models,because of its unique characteristics of decentralization,hard to tampering,trustless multi-party collaboration,and anonymous privacy,and has received widespread attention from industry and academia.Consensus algorithm is the core module of blockchain technology,which can support the orderly large-scale collaborative cooperation of blockchain nodes in a decentralized and trustless network environment,and its performance directly affects the overall efficiency of the blockchain system.Therefore,studying the challenges of the consensus algorithm in the blockchain technology faced in actual production,improving transaction processing capabilities,and expanding the size of the consensus node have important research significance and value.The classic distributed consistency algorithm PBFT(Practical Byzantine Fault Tolerance)and its variants are widely used in the consortium blockchain,but the following problems still exist.First,it cannot support node dynamic joining and exiting in the consensus process like most public blockchains.Second,in the consensus process,the space occupied by the voting signature message is too large.Third,the complexity of the message is exponential,and the transaction throughput rate is low,so it cannot support large-scale application.In this paper,a series of work has been carried out to solve the above problems,including:(1).Aiming at the problem that the PBFT algorithm does not support node changes in the consensus process,this paper proposes a solution based on the Byzantine Fault Tolerance consensus algorithm that supports hot swapping of nodes.This solution uses the way of weak centralized routing table to build the nodes status table,and support node blacklist function.Any node can dynamically join and exit the consensus network during the consensus process without restarting network.(2).In response to the problem that the voting signature message takes up too much space,this paper introduces a multi-signature method based on BLS,which can improve the threshold voting in the consensus process.The threshold voting for m-of-n scenarios is implemented,and the length of signatures and public keys are constant level,which effectively shortens the delay of threshold voting,and provides the underlying technical support for the threshold voting scenario in the consensus process and node dynamic hot swapping process.(3).To address the problem that the complexity of the message is exponential so cannot be applied on a large-scale network.This paper proposes a Flexible Byzantine Fault Tolerance(FBFT)algorithm based on the above improvement scheme.By improving the PBFT consensus process,the message complexity is reduced from(!)to(),so that the consensus algorithm can be applied to a larger-scale blockchain network.Experiment with the FBFT algorithm under the FISCO-BCOS platform,and result shows that compared with the PBFT algorithm,the transaction throughput rate is increased by 4.54%.Finally,based on the FBFT consensus algorithm proposed in this paper,we design and implement a blockchain account data sharing platform to achieve account information sharing among enterprise alliances.Then,we constructed a medical joint account sharing platform in the consortium blockchain scenario,which through the case of the transfer operation of patients' medical records data between hospitals to verify the effectiveness of the FBFT algorithm.