Resource Trading Incentive Mechanism Based on Stackelberg Game and Auction Theory in Computing Power Network

Computing power network (CPN) is a new architecture proposed in recent years. It has integrated multidimensional resources, such as computing, network and storage resources in cloud, edge and end to realize more flexible resource allocation and task scheduling. Consumers in CPN want to purchase some resources to meet their service demands and gain some profits, while operators would like to obtain payments by selling their idle resources to others. However, consumers usually prefer to purchase resources from large and public cloud operators in reality, while small and private operators who lack the recognition struggle to obtain the opportunity to join in the sales. Therefore, both parties may lack the passion and enthusiasm to participate in CPN transaction and there is a great need for a resource trading incentive mechanism. In this paper, we propose a three-tier joint optimization framework to increase their motivations and establish a centralized computing service platform (CSP). We use the Stackelberg game to describe the resource-purchasing process of computing service demanders (CSDs) and use the auction theory to describe the task-scheduling process of computing service operators (CSOs). In this way, both parties can gain their profits so that they will have a greater willingness to participate in the transaction. The simulation results demonstrate the feasibility and superiority of our scheme.