Research On Cluster Management And Resource Allocation In Downlink Non-orthogonal Multiple Access System

With the evolution of 5G network,the Internet and the Internet of Things have emerged diversified service scenarios,while the existing traditional communication technology has been difficult to meet people’s requirements for massive connections and high spectrum efficiency in diversified business scenarios.The traditional multiple access method is orthogonal multiple access,although it can avoid the same frequency interference between signals,but with the exponential increase of the number of access users,bandwidth resources tend to be limited,the system throughput is difficult to improve the space.Non-orthogonal Multiple Access(NOMA)technology has attracted much attention because of its great contribution to improving the utilization rate of system spectrum resources.NOMA takes advantage of the differences in channel states between users.At the sending end,multiple users can reuse the same spectrum resource block for information transmission in the form of overlay coding,different users share the same spectrum resource in nonorthogonal form.This non-orthogonal multiple access can not only maximize the utilization of spectrum resource,but also significantly increase the system throughput and user access number.This paper focuses on the following two issues:1.User Clustering:Based on the existing research,a hybrid user clustering method based on user channel gain difference is proposed.In this scenario,the number of cell edge users will be far more than the number of users,if the traditional way of near-far user pairing will make part of the edge users within the same cluster,result in the same frequency interference between the users,so this article makes a center block edge users and multiple users share the same resources,and the edge between the user bandwidth is independent of each other,In this way,the same frequency interference between edge users can be reduced and the resource utilization can be maximized.The simulation results comprehensively compare several existing clustering methods,and in the scenario presented in this paper,the proposed scheme performs best.2.Power Allocation:First analyzed the hybrid cluster center user power allocation factor within scope,given by minimum difference between the receiving end user signal values of ceiling,target rate by the receiving end of the lower limit of the minimum value is given,followed by interval iterative search way,with the aim to optimize the system throughput,combined with the rate at which the end user and fairness,The power distribution factor among edge users is given according to fractional power distribution.Finally,through the numerical simulation experiment of the system platform,it is further verified that the proposed power distribution scheme can significantly improve the system throughput compared with the traditional scheme.