Research On Key Technologies Of MIMO-NOMA System

The core idea of Non-Orthogonal Multiple Access technology is to provide services to multiple users in the same resource block.However,due to the particularity of the serial interference cancellation decoding at the NOMA receiving end,all users through the NOMA service will make the receiving end highly complex and decoding errors greatly increased.Therefore,users are usually clustering/pairing,and the clustering/pairing users are served by NOMA.The optimal pairing/clustering can be obtained by exhaustiveness,but obviously this is not realistic when the number of users is large,so it is expected to find a better clustering/pairing scheme under the condition of compromise between complexity and performance.Secondly,when users are clustering/pairing successfully,due to the principle of NOMA,users with poor channel gain are allocated more power,and users with strong channel gain are allocated less power,so reasonable power allocation is crucial to the performance of the system.Based on the above problems,this thesis does the following research from user pairing and power allocation:Firstly,when the base station has a single antenna,a new user pairing algorithm is proposed,which can adaptively select the optimal group for pairing according to the number of users.The users are sorted based on the strength of the user channel gain,and then the users are divided into various even groups according to the number of users.Find the optimal even-group pairing combination.Secondly,in order to further improve the throughput of the system,a power allocation that maximizes the sum rate is proposed.Under the constraints of minimum user target rate,NOMA principle and base station power,the sum rate objective function of two users is formed.For this optimization problem,since it is assumed that the pairing has been successful,the solution of the objective function is obtained directly by the Lagrangian method.However,since only the power allocation of NOMA users is considered and user pairing is not considered,the obtained solution is a suboptimal solution.Finally,this thesis studies the NOMA user pairing problem of multi-antenna base station.Different from traditional research points,this thesis assumes that users are distributed in a regular hexagon and obey Poisson distribution.Secondly,the improved K-means algorithm is used in this model to cluster users.The advantage of clustering is that users within a cluster have high correlation,while users between clusters have low correlation,so that the interference between users is small.After the clusters are divided,users are paired with the maximum and minimum channel gains within the cluster.Since there is interference between unpaired users,in order to eliminate the inter-pair interference,this thesis adopt the precoding technique of singular value decomposition to eliminate the interference.Based on the above research,it is proved that when the base station has a single antenna,the proposed user pairing algorithm has advantages in increasing the number of user pairs and system throughput.Secondly,under the same conditions,the proposed maximization and rate power allocation can achieve better system performance.When the base station has multiple antennas,the improved K-means pairing algorithm based on the Poisson distribution model can achieve better system rate than existing algorithms,and it is found that with the increase of the number of transmitting antennas,its performance is better.