Research On Performance Models Of Blockchain Systems Under Typical Consensus Mechanisms

Blockchain is a new computer technology with a brand-new decentralized structure,which is gradually emerging with the increasing popularity of digital encryption currencies such as Bitcoin.It has attracted the attention of various countries and major organizations around the world.Blockchain has a very wide range of application scenarios in the fields of medical care,finance,Internet of things,transportation because of their decentralization,non-tampering and anonymity.The consensus mechanism directly affects the consensus process of blockchain and the way of block generation,and further affects the performance of the blockchain.This dissertation establishes a queuing model for blockchain systems by comprehensively applying theories such as stochastic processes,matrix geometric solutions in several commonly used consensus mechanisms(such as the Practical Byzantine Fault Tolerant consensus mechanism in the consortium blockchain,the Proof of Stake consensus mechanism and the Delegated Proof of Stake consensus mechanism in the public blockchain and the Raft consensus mechanism in the private blockchain).What’s more,the steady-state probability of the system is calculated respectively,and the performance of the blockchain is evaluated and analyzed.At the same time,applying the blockchain consensus mechanism to smart healthcare systems effectively improves medical information security.The main results of this dissertation are as follows:(1)The performance model of blockchain systems under the Practical Byzantine Fault Tolerant(PBFT)consensus mechanism is studied.For the consensus process of the PBFT consensus mechanism,a level independent quasi-birth-and-death-process is established.By using the matrix geometric solution,the necessary and sufficient conditions for the steady-state of the PBFT consensus mechanism in the blockchain are obtained.By using the steady-state probability vector of the system is calculated,so as to realize the analysis of the steady-state probability of the PBFT consensus mechanism in the blockchain,and then evaluate the performance of the PBFT consensus mechanism blockchain system.This dissertation provides a complete analysis method for studying the mainstream consensus mechanism in the blockchain system alliance chain.(2)The performance model of blockchain systems under the Proof of Stake(Po S)consensus mechanism is studied.The Po S consensus mechanism is analyzed by using Markov models with absorption states and summarized into two situations.The first situation is when there is only one vote to verify the transaction package,the board of directors successfully verifies the transaction package and adds it to the chain(or fails to become an orphan block);The second is that after the transaction package is successfully verified by continuous voting,the transaction package generates a new block until multiple blocks are verified by continuous voting.For these two situations,this dissertation calculates the first passage time required for the board of directors to conduct one voting process(successful or failed voting verification)and continuously vote until multiple blocks are generated.Based on this,a blockchain system with a Po S consensus mechanism is constructed.The steady-state conditions of the system are given,by using the matrix geometric solution method,what’s more,the steady-state probability is calculated,and expressions for some system performance indicators are given to analyze and evaluate the performance of the Po S consensus mechanism in blockchain systems.This dissertation develops a calculation method for a stochastic model of consensus mechanism in blockchain systems.(3)The performance model of blockchain systems under the Delegated Proof of Stake(DPo S)consensus mechanism is studied.A three-dimensional continuous-time Markov process with absorption state is used to calculate the first passage time of witness nodes for continuous voting verification of transaction packages until multiple blocks are generated,and the first passage time obeys a continuous-time PH distribution.In addition,the DPo S blockchain systems in which the witness node elects multiple blocks are established,and the expressions of the steady-state conditions and performance indicators of the system are obtained by using the matrix geometric solution method.What’s more,the accuracy of the theoretical results is verified by a numerical example.This dissertation provides a computational basis and valuable reference for studying the consensus mechanism of DPo S in blockchain systems.(4)The performance model of blockchain systems under the Raft consensus mechanism is studied.The first passage time of a Leader node is calculated by using a four-dimensional continuous-time Markov process with absorption state.The first passage time for Candidate nodes in the system to successfully verify voting and chain up(or fail)is calculated,by establishing two Markov processes with absorption state finite states respectively.The blockchain system of Raft consensus mechanism is studied by establishing an M(?)PH~b/PH~b/1type queuing model.What’s more,the steady-state conditions of the system is obtained by using matrix geometric solution,and the steady-state probability vector distribution is established.In addition,under the condition of system steady state,the performance of Raft consensus mechanism of blockchain system is evaluated and the key performance indicators are given.This dissertation provides practical and feasible solutions and computational method support for studying the Raft consensus mechanism in blockchain systems.(5)Empirical analysis of smart healthcare systems based on blockchain.Taking the smart medical system of Beijing CG Hospital as the research object,and based on the operational characteristics of the smart medical system of Beijing CG Hospital,a solution for medical information storage and sharing is provided using blockchain technology.By utilizing the PBFT consensus mechanism and Po S consensus mechanism in the two blockchains of‘the consortium blockchain+the public blockchain’,the storage chain and sharing chain of Beijing CG Hospital’s smart healthcare system are established,and a random model of the smart healthcare system is constructed;Through empirical analysis,it has been proven that the blockchain consensus mechanism can greatly improve system performance and stability,making the management of medical information more efficient,stable,and secure.