Incentive Mechanism for Edge Computing-based Blockchain : A Sequential Game Approach

Abstract

The development of the blockchain framework is able to provide feasible solutions for a wide range of Industrial Internet of Things (IIoT) applications. While the IIoT devices are usually resource-limited, how to make sure the acquisition of computational resources and participation of the devices will be the driving force to realize blockchain. In this work, an edge computing-based blockchain framework is considered, where multiple edge service providers (ESPs) can provide computational resources to the IIoT devices. We focus on investigating the trading between the devices and ESPs, where ESPs are the sellers and devices are the buyers. A sequential game is formulated and by exploring the sequential Nash equilibrium, the existence of the optimal solutions of trading strategies can be proved. Then, a deep Q-network-based algorithm with a modified experience replay update is applied to find the optimal strategies. Through simulations, we demonstrate the effectiveness of the proposed incentive mechanism.