A Game for Task Offloading in Reputation-Based Consortium Blockchain Networks

Abstract

This letter studies the task offloading in reputation-based consortium blockchain networks, where the tasks are transmitted to Edge Computing Servers (ECSs) due to limited resources. We propose a novel Delegated Proof of Stake (DPoS) consensus mechanism in which validation nodes (including active nodes and backup nodes) are voted based on their reputation. The incentive identifies the reputation and the consensus delay as two major factors determining the reward. A three-stage Stackelberg game is developed to jointly minimize cost of the users and maximize utilities of the master node and the validation nodes. We analyze the unique Stackelberg equilibrium exists in the proposed game by the backward induction. The simulation results demonstrate that the designed incentive is feasible for trust management, and the proposed consensus has lower delay and higher decentralization compared with the traditional DPoS.