Research On Precoding Techniques In Large-scale MIMO Systems

Massive Multi Input Multiple Output(MIMO)technology,as one of the key technologies of 5th generation(5G)mobile communication,will be applied more and more widely.Pilot contamination is the main factor that affects the capacity of the massive MIMO system,and it is also the main bottleneck affecting the business of large-scale MIMO systems.The system precoding is a very effective way to reduce pilot contamination in large-scale MIMO system.At the same time,it can reduce the complexity of signal detection for the receiver.This paper have a more in-depth study about precoding technology,aiming at the problem of high complexity of the existing linear precoding,research on precoding algorithm based on iterative method,ensure that the implementation complexity of the precoding algorithm is reduced under the condition that the system performance is not affected.The main contents in this paper are as follows:Firstly,the basic processes of large-scale MIMO channel modeling and the channel asymptotic orthogonality are briefly introduced in this paper,and the impact of antenna spatial correlation on the channel capacity in large-scale MIMO system is analysed,through the channel characteristic simulation.Based on least square(LS)and minimum mean square error(MMSE)channel estimation,In this paper,we analyses the causes and effect of pilot contamination,gives theoretical derivation of the effect on the system capacity.Secondly,the system precoding technology for large-scale MIMO is mainly studied.The performance of linear precoding algorithm in large scale MIMO system is simply analyzed.Because the matrix inversion causes high complexity,the overrelaxation precoding algorithm is studied,however the successive over relaxation(SOR)algorithm is sensitive to the relaxation factor,so a weighted symmetric successive over relaxation(W-SSOR)precoding is proposed.The simulation results show that the proposed algorithm not only reduces the sensitivity to the relaxation factor,but also has a certain improving convergence speed.Finally,in order to reduce the algorithm complexity,the iterative precoding algorithm is studied more deeply in this paper,aiming at the slow convergence speed of the existing conjugate gradient(CG)precoding,a new CG iterative algorithm based on SSOR pre-treatment is proposed,and the iterative scheme is improved.The simulation results show that the proposed algorithm not only have a good control of complexity,also has greatly improved the convergence speed.