Stackelberg Game Based Edge Computing Resource Management for Mobile Blockchain

Abstract

Blockchain can provide a dependable environment for mobile applications. As the core of blockchain, mining consumes a large amount of computing resources, while mobile devices are too resource-limited to perform the mining. Offloading mining computation tasks to an edge computing service provider (ESP) or a cloud computing service provider (CSP) is considered as a feasible solution to mobile blockchain mining. However, the computing resources of the ESP are not unlimited. Therefore, rational edge computing resource management is critical to maximizing the utilities of the ESP and the miners. Most of the existing work assumes that the service provider is an ESP or a CSP, or both. In this paper, we construct a computing offloading model consisting of multiple miners, an ESP, and a CSP, where the ESP and the CSP are independent of each other. We formulate a Stackelberg game with the ESP as the leader and the miners as the followers for optimal pricing-based edge computing resource management. We analyze the existence and uniqueness of Stackelberg game equilibrium and derive the miners' optimal computing resource requests. We then propose an effective golden section based Stackelberg game equilibrium searching algorithm (SES) for resource pricing. We conduct the experiments through simulations. The simulation results show that the proposed computing offloading model and algorithm can achieve high unit service utilities of both the ESP and the end devices.