An improved PBFT consensus algorithm based on grouping and credit grading

Abstract To address the problems of high communication complexity, the random selection of master nodes, and limited supported network size of the Practical Byzantine Fault Tolerance (PBFT) consensus algorithm for consortium chains, an improved Byzantine Fault Tolerance (CBFT) algorithm based on grouping and credit hierarchy is proposed to optimize the nodes of large-scale consortium chains structure. First, the network nodes are divided into different consensus sets according to their response speed to the management nodes, and the consensus is carried out inside and outside the group respectively; based on this, a credit grading mechanism is proposed and a credit calculation formula is introduced to select the management nodes; finally, a simulation and performance testing system based on this improved scheme is built. The experimental results show that CBFT has less communication overhead, shorter latency, and higher throughput than PBFT, and it is more obvious with the increase in the number of nodes, which meets the needs of large consortium chains.