Research And Optimization Of User Grouping Scheme And Power Allocation Algorithm In Downlink NOMA

How to realize the access of a large number of devices is a key difficulty in 5G research.Non-orthogonal multiple access(NOMA)technology,as an important candidate for 5G,exploits the power domain so that multiple user signals can be superimposed on the same time-frequency resource block for transmission,which greatly improves the system capacity and spectrum efficiency.At the same time,a new problem is introduced,that is,how to allocate resources to superimposed users.The problem of resource allocation can be divided into two,one is user grouping,and the other is power allocation.However,most existing user grouping algorithms and power allocation algorithms have the following problems:high computational complexity,poor system performance,bad user fairness and lack of expandability.Motivated by the above problems,this thesis has carried out research on the resource allocation problem on downlink NOMA.In order to solve the problem of user grouping,this thesis proposes a user grouping algorithm based on variable neighborhood search.The variable neighborhood search algorithm is a heuristic search algorithm that can find the optimal or suboptimal solution to the problem in a short time.Due to the use of multiple neighborhood structures to search alternately,the algorithm achieves a better balance between concentration and evacuation and has better global optimization capabilities.Simulation proves that the algorithm proposed in this thesis has better performance in system throughput than the typical existing user grouping algorithms.Aiming at the problem of power allocation,this thesis proposes a power allocation algorithm based on bacterial foraging optimization.The search procedure is optimized by imitating the foraging behavior of bacterial,and the performance of the approximate full-space search algorithm is achieved under the premise of greatly reducing the computational cost.As the geometric mean throughput is used as the fitness function,the algorithm achieves a good balance between system throughput and user fairness.Finally,the effectiveness of the algorithm proposed in this thesis is verified by simulation comparison.The user grouping algorithm and power allocation algorithm proposed in this thesis have no limitation on the number of superimposed users,and can be generalized to the multi-user multiplexing situation without no more effort and can adapt to more application scenarios.