Research On Incentive Strategy Based Resource Optimization Algorithm Of Fog Computing Networks

The fog computing network has attracted extensive attention from scholars at home and abroad due to its low latency,real-time performance,and distributed computing.In order to make resource utilization more reasonable,resource optimization algorithms in fog computing networks have become one of the research hotspots.The thesis mainly studies resource optimization algorithms in fog computing networks,and uses game theory and other theoretical tools to allocate computing resources,storage resources,and energy resources,which improves resource utilization.The main contents of the thesis is as follows:1)Aiming at the problem of resource optimization between fog nodes in fog computing networks,a task offloading incentive algorithm based on the ascending-bid auction mechanism is proposed to encourage fog nodes to share their computing resources.In this algorithm,the fog controller acts as an auctioneer,and the fog nodes act as bidders and feedback the amount of data they are willing to process.The auction process iterates until the requirements of the auctioneer are met.Simulation results show that compared with cloud data processing,the proposed algorithm not only has low energy consumption,but also greatly improves the quality of network services.Aiming at the resource optimization problem of fog nodes in fog computing networks,a computing resource allocation algorithm based on DQN(Deep Q-learning Network)is proposed.The algorithm first gives the definition of task priority and sets the environment,state,actions and rewards,and then builds a reinforcement learning model for resource allocation.Finally,the simulation results show that the average task execution rate of the algorithm is much higher than the traditional computing resource allocation algorithms.2)Aiming at the optimization of storage resources in fog computing networks,a storage resource incentive optimization algorithm based on Stackelberg game and noncooperative game is proposed.In this algorithm,the fog control node acts as a leader to incentivize fog nodes,and the fog nodes as follower share their own storage resources.We obtain the Nash equilibrium through optimization.Simulation results verify the effectiveness and convergence of the algorithm.3)Aiming at the problem of energy optimization in fog computing networks,a SMETO(Stable Matching for Energy-Minimized Task Offloading)algorithm based on matching theory is proposed.In this algorithm,task nodes can be matched to helper nodes that are willing to process data for them with the lowest energy consumption,and helper nodes can also be matched to the ideal task nodes.Simulation results show that compared with the random task offloading algorithm,the proposed algorithm can reduce energy consumption by 67%.