论文部分内容阅读
Blockchain with these character-istics of decentralized structure, transparent and credible, time-series and immutability, has been considering as a promising technol-ogy. Consensus algorithm as one of the core techniques of blockchain directly affects the scalability of blockchain systems. Existing probabilistic finality blockchain consensus al-gorithms such as PoW, PoS, suffer from power consumptions and low efficiency; while abso-lute finality blockchain consensus algorithms such as PBFT, HoneyBadgerBFT, could not meet the scalability requirement in a large-scale network. In this paper, we propose a nov-el optimized practical Byzantine fault toler-ance consensus algorithm based on EigenTrust model, namely T-PBFT, which is a multi-stage consensus algorithm. It evaluates node trust by the transactions between nodes so that the high quality of nodes in the network will be selected to construct a consensus group. To reduce the probability of view change, we pro-pose to replace a single primary node with a primary group. By group signature and mutual supervision, we can enhance the robustness of the primary group further. Finally, we an-alyze T-PBFT and compare it with the other Byzantine fault tolerant consensus algorithms. Theoretical analysis shows that our T-PBFT can optimize the Byzantine fault-tolerant rate, reduce the probability of view change and communication complexity.