Research on PBFT Consensus Algorithm Based on Reward and Punishment Mechanism

Abstract

In the blockchain network constructed by the Practical Byzantine Fault Tolerance (PBFT), there are problems that the master node needs to be replaced frequently when the master node and the sub-node break down, and the same information is verified multiple times when the sub-node does not work normally. In response to these problems, the award and penalty mechanism is applied to the PBFT consensus algorithm and supervision nodes are introduced to supervise the master node and the sub-node in the blockchain network. Construct a tripartite evolutionary game model with the master node, suspected sub-node and supervision node as the main body to analyze the changes of the three in reaching the three stable states of normal information transmission, normal information verification, and supervision of other nodes before and after the improvement of the PBFT consensus algorithm, and use Matlab software for simulation calculation. The research result proves that the application of the award and penalty mechanism to the PBFT algorithm can improve the efficiency of the master node and the suspected sub-node in normal information transmission and normal information verification in the blockchain network. And it can also increase the participation proactivity of the master node and the suspected sub-node in information verification and the running for the master node. At the same time, increasing the system reward and the reward value, penalty value, and punishment credit value based on the award and penalty mechanism will also improve their efficiency of reaching normal information transmission and normal information verification, and ultimately ensures that the constructed blockchain network can efficiently achieve the goals of information verification and active participation of nodes in verification.