Research On Resource Allocation For Network Slices In Fog Radio Access Networks

In order to meet the differential demands in a variety of application scenarios,the network slicing is proposed for the next generation mobile communication.The traditional physical network is virtualized and divided into several customized network slices,which serve different applications.The resource allocation is a key issue to ensure successful working,so the method of resource allocation for network slicing in the fog radio access network is studied in this paper.Firstly,in order to cope with many problems caused by traditional network wireless resource management,which can not adapt to the characteristics of network slicing,a unified resource management framework for network slicing is proposed,which is a two-layer structure and can effectively alleviate the computational pressure of the centralized distribution.Meanwhile,the distributed interfaces allow resource exchange between slices,which enhances the flexibility of resource reallocation.Finally,this framework can improve the efficiency of resource management for the whole system.Secondly,due to the opening of network capability of network nodes,new types of resource will appear in the access network.In order to solve the resource allocation problem in the access network,a method of joint allocation for radio resources and computing resources is proposed to provide better services for users,and an optimization problem aiming at minimizing energy consumption is established.In order to solve the problem with nonlinear optimization function and coupling variables,the proposed algorithm uses Lagrange partial relaxation to decouple the original problem into several sub-problems,which can be solved respectively,and achieve the optimal allocation scheme..Simulation results show that the joint resource allocation of radio and computing resources can effectively improve the overall performance of network slices comparing with the traditional single resource optimal allocation.Thirdly,the network needs to meet different performance requirements at the same time in the coexistence scenarios of different types of network slices.A specific application scenario with high rate slice and low latency slice is studied,and an average delay optimization problem for one slice with user data rate guarantee for the other slice is formulated.To solve this nonconvex integer optimization problem,the proposed algorithm splits the original problem into several sub-problems,which can be solved by Hungarian algorithm,linear integer programming and geometric programming,respectively,and obtains the final result by loop iteration.Simulation results show a fast convergence speed,and the local performance of the proposed algorithm is close to the optimal performance of the exhaustive search algorithm.In this paper,the resource allocation for network slicing based on different application scenarios is studied,which provides the theoretical basis for joint resource allocation and differential application performance.It also provides a practical guarantee for the business deployment of network slices.