Research On Multi-user Technology In 5G Communication System

The fifth-generation(5G)wireless communication system is designed to meet and solve various communication technical requirements and challenges,and can greatly improve the data transmission rate,system capacity and energy efficiency.As one of the key technologies of 5G,massive MIMO technology can provide higher data rates without increasing additional bandwidth in wireless communication services.Combining beamforming technology with massive MIMO can effectively compensate for the path loss of data transmission.Moreover,in a massive MIMO system with a large number of users,in addition to beamforming,user selection technology that improves system performance by optimizing resource allocation is also indispensable.This paper has studied a multi-user transmission scheme suitable for massive MIMO systems,aiming to solve the problems of large pilot overhead and computational complexity in existing beamforming and user selection algorithms.This paper has proposed two solutions,one is the decoupling design of beamforming and user selection,and the other is the joint design of beamforming and user selection.The main research content and work results of this article are as follows:1.Based on high-precision positioning technology and using position information to obtain channel state information(CSI),this paper has proposed a two-stage beamforming scheme:select the outer-beamformer from the antenna steering vector set;based on the equivalent channel matrix,use Iterative QR decomposition algorithm realizes inner-beamformer.Based on the random matrix theory(RMT),this paper has derived the user's asymptotic SINR.The proposed user selection algorithm includes two stages: selecting semi-orthogonal users to eliminate inter-user interference;selecting users with larger asymptotic SINR to ensure channel energy.2.In the joint design of beamforming and user selection,this paper has proposed a novel optimization framework.By expressing the joint optimization variables in the high-dimensional space,the maximization of the weighting sum-spectral efficiency can be expressed as the maximization of the Rayleigh quotient product.Although this is still a non-convex problem,this article has proposed a computationally efficient algorithm called Generalized Power Iterative Beamforming(GPIB).The algorithm will converge to the fixed point(local maximum)of the objective function,thereby ensuring the first-order optimality of the solution.By adjusting the weights of the weighting sum-spectral efficiency,GPIB can provide a joint solution for user selection,power allocation,and downlink beamforming.