RG-PBFT: A Consensus Protocol based on Reputation and Grouping Mechanisms for Alliance Blockchain

In the overall network environment of the blockchain, consensus algorithms are designed to ensure the security, decentralization, and scalability. Currently, many alliance blockchains use the Practical Byzantine Fault Tolerance (PBFT) protocol to maintain system consistency. Since all nodes participate in consensus, the PBFT causes unacceptable system delay. In this paper, an improved PBFT protocol based on reputation and grouping mechanisms, named as RG-PBFT, is proposed to dynamically adjust the number of network nodes participating in consensus. A reputation performance score model will be introduced to reduce the communication frequency and response delay between nodes. In addition, the primary node selection strategy is implemented with a verifiable stochastic function (VRF). The resulted experiments show that with the increasing in the number of network nodes, the RG-PBFT protocol improves performance in terms of communication overhead, consensus efficiency and scalability.