Research On Resource Allocation Algorithm Based On Slicing In B5G Network

To deal with the development trends from the mobile Internet to the ubiquitous Internet of Things,the thousand-fold increase on data traffic,the geometric increase on the number of connected devices and varieties on the service requirements,virtualization technology,network slicing and non-orthogonal multiple access(NOMA)techniques are introduced into the fifth generation mobile communication systems to satisfy the diverged service requirements,the service scalability and service diversity.With the introduction of a series of new techniques,how to realize the effective utilization of resource under the service requirement and the limited resource constraints becomes a challenging problem.Under the circumstance of networking slice and NOMA technique utilized,this thesis mainly explores the solution to the resource scheduling problem for the users inside and across the network slicing to optimize the efficiency of resource utilization.The main work is listed as follows:(1)The uplink joint resource optimization problem on the radio and computing resource with NOMA is analyzed.Due to different performance requirements of each slicing,the utility functions for different 5G network slicings are formulated,and then the total system optimization objective is established by integrating each slicing utility function together.However,the formulated optimization problem is a NP-hard problem,a step-by-step solution based on geometric programming and Hungarian algorithm by decomposing the original problem into multiple sub-problems is proposed to obtain an approximately optimal resource allocation.(2)The performance of the proposed scheme is validated through a Monte-carlo simulation platform using MATLAB.Performance comparisons are done in the aspects of data rate,throughput,latency and quality of service between the proposed and traditional schemes.The verification results show that the proposed scheme is able to increase the system throughput and spectrum efficiency while satisfying the corresponding service requirements,which helps to increase the system capacity and meet the application demands.