A Novel Distributed Computing Resource Operation Mechanism for Edge Computing

Traditional cloud computing services are generally deployed centrally, which can not meet the ultra-low delay requirements of services such as VR/AR, Internet of Vehicles, and cloud games. Edge computing can solve this problem by deploying nodes near users. Edge nodes are geographically dispersed and great in number, which usually are not provided by single edge computing service provider but by multiple. These providers are in competitive relationship. Tradition centralized cloud computing resource operation model can’t achieve unified resource orchestration and operation among edge nodes. Therefore, a fair and efficient computing resources operation mechanism is necessary. In this paper, firstly, we design a new distributed computing resource operation strategy. In the state of perfect competition, each edge node decides the price of local computing resources by itself. Users can freely choose among many providers according to their own strategies, realizing fair competition among different edge computing service providers. Then, to increase the market size while maximize the revenue of edge computing service providers, we define a multi-objective optimization model. Finally, we design a dynamic price adjustment mechanism to calculate reference price for edge node. The simulation results show that the computing resource operation mechanism we designed is significantly better than the traditional edge computing model.