Research On Resource Allocation In NOMA System

Non-orthogonal Multiple Access,which is considered as one of the candidate techonlogies of 5G due to the properties of improving spectrum utilization,can increase the number of access users and is compatible with existing communication architectures.The main idea of NOMA is that different users share the same spectrum resources at the same time,and the users are distinguished by power domain multiplexing at the receiver.Therefore,the research of resource allocation in NOMA system is promising.The existing user clustering algorithm is achieved by setting the threshold value of channel similarity,which leads to a certain randomness due to the difference of threshold.Therefore,in this thesis,a user clustering algorithm using symmetric matrix is proposed,wherein a nonlinear transformation of channel similarity and users' channel gain-difference are construced.Furthermore,all the solutions of user groups can be obtained by utilizing symmetry characteristic of matrix,so the randomness due to setting the correlation threshold is avoided.The simulation results show that the proposed user clustering algorithm can improve the system capacity by 5.51%-21.80%,3.98%-15.30% and 1.56%-6.69% compared with the traditional algorithm when the threshold is 0.75,0.85 and 0.95,respectively.In view of the existing power allocation mechanism used in NOMA,some problems are existed,such as the power allocation factor is fixed,the complexity is too high,the power allocation factor needs to be obtained through simulation analysis,or the power allocation increases the users' interference.Therefore,in this thesis,a minimum power allocation mechanism in NOMA is proposed,and transform the problem of minimizing the transmission power to meet the users' target rate into investigating the users' group precoding vector,then the minimum power is allocated between the two users in one cluster,so that the system capacity can be maximized under the satisfaction of users' quality of service and the limitation of NOMA,and the problem can be solved by the Lagrangian multiplier method.The simulation results show that the proposed minimum power allocation mechanism can reduce the transmission power,meanwhile,improve the system capacity.Aiming at the problem that some user clusters in the base station can not access successfully due to the excessive number of users and limited spectrum resources,a mechanism using energy detection spectrum sensing is proposed,so as spectrum borrowed from adjacent base station can be allocated,and the frequent information interaction among base stations can be avoided,and the spectrum sensing technology which can adaptively detect the idle spectrum and generate the avaliable spectrum set.Then the borrowed spectrum is allocated by using matrix,and the limitation in the existing spectrum borrowing mechanism which allocating idle spectrum by taking utility values in descending order is avoided.The simulation results show that the mechanism proposed can not only improve the system capacity and spectrum link efficiency,but also reduce the degree of unsatisfied users.