Multi-layer Resource Deployment Strategy For Mobile Edge Computing

With the rapid development of mobile Internet and artificial intelligence,all kinds of mobile terminal applications are becoming larger and larger,and the demand for computing resources is also increasing.Due to the limitation of equipment and volume of mobile terminal,its computing power and energy consumption can not support the application running locally.Traditional cloud computing paradigm usually has high transmission delay when processing computing tasks,so it can not meet the delay requirements of delay sensitive applications.Therefore,the researchers proposed mobile edge computing,whose core idea is to sink part of the cloud computing processing power to the network edge closer to the user,so as to solve the above problems.Obviously,the premise of using mobile edge computing technology is to deploy certain computing resources in the network edge nodes.Therefore,the resource deployment in mobile edge computing network is very important.In this paper,the edge computing resource deployment problem under flat and hierarchical network structure is studied systematically.The goal is to improve the effectiveness of resource deployment and improve the efficiency of the overall edge computing system.Flat network structure is a network model based on wireless metropolitan area networks and Ad Hoc network.Each network node is connected according to a certain topology shape to form a peer-to-peer structure.Based on the flat network model,the model of resource deployment and task scheduling is established.Resource deployment determines where and how many resources are deployed,and task scheduling allocates random arriving user tasks to corresponding resources in real time.Based on the above model,a joint optimization problem of resource deployment and task scheduling is proposed.The goal of the problem is to maximize the system benefits with the minimum deployment cost on the premise of ensuring the quality of user service.The analysis shows that the problem is a NP-hard nonlinear optimization problem.Through the equivalent transformation of the optimization problem,the original problem is transformed into an equivalent linear programming problem,so that the optimal deployment strategy can be solved by conventional algorithm.At the same time,three suboptimal deployment strategies with low complexity are proposed.The simulation results show that the three suboptimal deployment strategies can achieve good system performance.The hierarchical network structure is mainly based on the cellular wireless access network represented by 5G.Its network node connection forms a multi-layer tree structure according to different functions.For example,the base station as a leaf node directly provides wireless access for users,the sink node as the penultimate root node gathers multiple leaf nodes,and the core node acts as the last level root node.The node gathers multiple base station nodes.Based on the hierarchical network model,the model of resource deployment and task scheduling is established,and two different mechanisms are considered,that is,sharing resources between nodes at the same level and nodes at the same level cannot share resources.Based on the above model,the joint optimization problem is constructed and the optimal strategy is solved by branch and bound method.The simulation results show that the hierarchical deployment strategy with sharing mechanism can achieve higher system benefit and lower delay cost with lower deployment cost,and the system stability is stronger.