Research On Dynamic Resource Management In Fog Computing

Cloud computing provides end users with on-demand computing services anytime,anywhere by centralized sharing of resources.In the cloud computing mode,the end user sends the computing requirements to the cloud computing through the network to obtain services.With the development of the Internet of things and the wide use of intelligent devices,a large number of end users and computing requirements make cloud computing face great challenges in terms of bandwidth,delay,reliability,and security.As a new computing mode,the computing resources of fog computing are distributed on the fog nodes at the edge of the network,which makes the terminal data can be processed at the edge of the network,reduces the service delay,and avoids the transmission of a large number of data in the network.At the same time,fog computing enables the network to connect massive terminal devices,which promotes the rapid development of the Internet of things.Compared with the data center providing massive resources in cloud computing,the resources of fog nodes in fog computing are limited,and a large number of terminal devices generate a large number of service requirements and computing data.How to effectively manage the resources of fog computing has become the focus of the industry.Aiming at the problem of resource management in fog computing,this paper studies the optimization of resource management to ensure the long-term benefits of the network in the scenario of multiple terminal devices connected to fog nodes.The main work and innovation of this paper are as follows:(1)Without considering the cooperative relationship between the fog nodes,for the scenario where multiple terminal devices are connected to a fog node,an optimization model with long-term average response time as the target and long-term average constraints is established.By analyzing the problem,Lyapunov optimization is applied to convert the model problem into an upper bound problem.Then,a distributed algorithm based on branch and bound is proposed.The terminal only needs to interact with the fog node with less information to obtain the resource allocation and offloading proportion.Finally,the boundary between the target value obtained by the algorithm and the ideal target value of the original model is analyzed theoretically,and the convergence of the algorithm is analyzed.The simulation results show that thedistributed algorithm based on branch and bound can achieve the required performance,and the algorithm is adaptive to time-varying traffic.(2)For the scenario where multiple fog nodes are connected to multiple terminals,considering the cooperation between fog nodes,an optimization model with long-term average energy consumption as the goal is established.Then,the Lyapunov stability migration method is used to transform the long-term average problem into a deterministic upper bound problem.Next,a distributed algorithm based on control variables is proposed.Each terminal device executes the proposed algorithm to obtain the optimal solution for computing resource allocation.Finally,it is verified through simulation that setting more fog nodes in the network can significantly reduce terminal equipment competition for resources and reduce terminal energy consumption.It is also verified that the system we build meets the compromise between time and energy.