Research On Optimization And Performance Evaluation Of Mainstream Consortium Consensus Algorithms

Blockchain technology has attracted increasing attention due to its decentralized,tamperproof,and traceable characteristics,and has been applied in many fields with broad prospects.As the core supporting technology of blockchain,consensus algorithm ensures the security and credibility of blockchain.Consortium blockchain sacrifices some degree of decentralization compared to public blockchain,but provides better security and privacy for data,and is conducive to regulation.It is currently the most promising type of blockchain for commercial applications.In recent years,the research on consensus algorithms suitable for consortium blockchain has grown rapidly,and more and more researchers have joined the improvement and research of consensus algorithms for consortium blockchain.However,there are currently a variety of consensus algorithms with different principles,and there is a lack of unified performance measurement standards and evaluation tools.At the same time,the performance of consortium blockchain consensus algorithms is far from that of traditional centralized storage structures,which hinders the development of consortium blockchain in large-scale application scenarios.In view of the current shortcomings of consortium blockchain consensus algorithms and the lack of effective evaluation standards,this paper selects the PBFT consensus algorithm as the research object,studies how to optimize the consensus algorithm,and then studies how to evaluate a consensus algorithm based on the optimized algorithm,in order to propose a performance evaluation framework for consortium blockchain.The proposed evaluation framework is verified and optimized by comparing it with other typical consensus algorithms.The main research work of this paper is as follows:(1)To address the inadequacies of the PBFT consensus algorithm,and a consortium blockchain Byzantine fault-tolerant algorithm based on a perfect binary tree communication topology(PBT-BFT)was proposed.To address the problem of insufficient security in the selection of primary nodes in the PBFT algorithm,a reputation evaluation model was designed to evaluate the behavior of nodes,and a reputation-based verifiable random function(R-VRF)was proposed.This ensures that the probability of random selection is positively correlated with the reputation value,thereby ensuring the fairness and randomness of the selection of nodes with different reputation values.Next,to address the high communication complexity of the PBFT algorithm,a perfect binary tree communication topology was designed to reduce the communication complexity to linear complexity.In addition,to further improve the performance,a rotating primary node and pipeline working mechanism were proposed to improve the efficiency of consensus.(2)Taking the mainstream consensus algorithms in consortium chains as the research object,a universal consensus algorithm performance evaluation framework is proposed.This framework quantifies the key indicators in performance evaluation,including throughput,latency,degree of decentralization,and consensus security level.Moreover,the framework has a configurable testing environment and features loose coupling and easy scalability,providing a unified testing platform for consensus algorithms.By evaluating mainstream consensus algorithms and the algorithm proposed in research content(1),the effectiveness of the framework is verified.(3)This paper conducts in-depth research on the optimization and evaluation of consensus algorithms in consortium blockchain.The performance bottlenecks,optimization strategies,and performance evaluation methods of mainstream consensus algorithms in consortium blockchain are analyzed,providing guidance for researchers to design new consensus algorithms.