User Clustering And Power Allocation For Uplink NOMA System

5G is considered as a mobile communication network system featuring high data rates,ultra-wide bandwidth,low latency and high-density access.Non-orthogonal multiple access has recently been considered as a key enabling technique for 5G cellular systems.There are several non-orthogonal multiple access technologies,such as power domain non-orthogonal multiple access NOMA,code domain non-orthogonal multiple access MUSA and SCMA.In NOMA,by exploiting the channel gain differences,multiple users are multiplexed into transmission power domain and then non-orthogonally scheduled for transmission on the same spectrum resources.Successive interference cancellation(SIC)is then applied at the receivers to decode the message signals.NOMA technology changes the power domain from single-user to multi-user for the first time,making the total wireless access and spectrum efficiency of the system effectively increased.For user clustering algorithm,this paper proposes a user clustering algorithm based on traversal search strategy to maximize channel quality difference among different users in the same NOMA cluster.The algorithm has the same number of users in each cluster,and on this basis,the clustering scheme that maximizes the average value of the channel quality difference among different users in each cluster can be obtained.By analyzing the complexity of the algorithm,we can find that the complexity of the algorithm in the indoor hot scene is completely within the controllable range.For the power allocation scheme,this paper first analyzes the system model with two users as an example.Combined with the research scenario,it uses an analysis method that limits the total transmit power in a cluster,and then introduces the power distribution factor,and then studies the power distribution factor.The impact of changes on system performance was considered in the study process.The constraints of different users' signal received power clarity and single user target data rate were considered.The results of the study showed that when two constraints are met,the users with better channel quality are allocated with more power.High transmit power increases system performance.Then the system model is generalized to the general situation.The throughput performance of the system is studied when the number of users in the cluster is an arbitrary constant,and the expressions of the power allocation factors of different users are given.Finally,the simulation results are analyzed,simulation results show that NOMA system can get better system throughput than OMA system.The last part of the dissertation builds a 5G system-level simulation platform based on C++ language in the windows system,and compares the throughput performance with the LTE system-level simulation platform under the same scene.In the specific implementation,the resource scheduling module is mainly designed.In this module,a user clustering algorithm and different user power allocations are implemented,and joint scheduling is performed for users in the same cluster.The results show that compared to LTE systems,NOMA systems can bring better throughput gains,especially when the number of users in the system is large.