Research On Delay-Aware Service Caching Mechanism In Mobile Edge Network

With the arrival of 5G era,Mobile telecom network service is transitioning from data center service mode to distributed Mobile edge computing(MEC)service mode.In this context,Network Function Virtualization(NFV)technology gradually attracted attention.Traditional Network infrastructure services,such as Network firewall,intrusion detection system and data transcoder,etc.,were used in Network Virtualization(NFV).Virtualization technology can be applied to deploy more flexibly to the general server of edge network in the form of Virtual Network Function(VNF),so as to reduce network service delay and improve user experience.At the same time,there are more and more network delay-sensitive applications such as VR application,AR application and network live broadcast.The network service providers that provide these services can also cache their services in the Data Center(DC)to the MEC network,so as to serve their users nearby.In view of the above two problems,this thesis studies the problem of service caching in MEC network,that is,how to deploy application services in edge network nodes and make more efficient use of MEC server resources.In this background,consider the following two issues background,one is only one application service provider,which provide basic network infrastructure operations of the business functions,at this time only need to consider how to satisfy the premise of application service delay,will be deployed to the user side of the edge server network service,makes the lowest cost of the overall application service cache,The other is that there are multiple application service providers,namely multiple application software service providers.In addition to the problem of how to deploy the application in the first case,the competition among application service providers for the edge server resources also needs to be considered.For the service placement problem with only one application service provider,this thesis establishes constraints on computing resources and bandwidth resources,constructs auxiliary graphs,unifies the modeling of service delay and transmission delay,and constructs a weighted path finding problem to constrain the maximum service delay.The goal is to model an optimization problem to minimize the cost of service deployment,and then describe the optimization problem as a combinatorial optimization problem(knapsack problem)and solve it using an approximate algorithm.For the multi-application service provider service cache problem,based on the first optimization problem,combined with Stackelberg game strategy,application service providers are divided into negotiable set and selfish set,and then the optimization problem is solved by applying the game theory,and finally the multi-application service provider service cache scheme is obtained.Finally,this thesis tests the performance of the two methods through simulation experiments and comprehensive experimental platform experiments,and the experimental results show that the algorithm proposed in this thesis is feasible.