A Modified Blockchain DPoS Consensus Algorithm Based on Anomaly Detection and Reward-Punishment

Abstract

The blockchain consensus algorithm can be leveraged to enhance IoT security and promote the efficiency of collaboration among IoT nodes. Delegated Proof of Stake (DPoS) can simultaneously meet the low-cost and high-efficiency requirements and improve the service quality of node collaboration. However, the malicious attacks, selfishness and insufficient enthusiasm of collaborative nodes can affect DPoS consensus process. In view of these challenges, we modify DPoS consensus algorithm, and design a self-selecting abnormal data detection algorithm to identify abnormal data from malicious attacks and selfishness behavior of nodes. Meanwhile, a game theory-based reward and punishment incentive mechanism is proposed to improve the voting enthusiasm of nodes. Simulation results show that the proposed mechanism can effectively reduce the effect of malicious attacks and nodes selfishness in DPoS consensus process.