Font Size: a A A

Research On 5G Massive MIMO Precoding Algorithm

Posted on:2020-01-26Degree:MasterType:Thesis
Country:ChinaCandidate:W H BaiFull Text:PDF
GTID:2428330602950656Subject:Engineering
Abstract/Summary:PDF Full Text Request
In the 5th generation(5G)mobile communication system,massive Multiple Input Multiple Output(MIMO)technology,as one of the core technique of its development,has greatly improved the spectrum efficiency,and the width of the transmission beam is gradually narrowed due to the surge in the number of antennas at the emission end.Therefore,the interference between different beams decreases significantly.Although this technology sufficiently enhance the performance,it requires high level of computation of system beam management and control,regarding the fact that the processing capability of existing transceivers cannot withstand large-scale beamforming,the investigation of high-performance,low-complexity precoding or detection algorithms have become the current research hotspots.In this paper,the signal processing problem of precoding technology in massive MIMO system was studied.The study of the mechanism was twofold,i.e.the selection of the precoding algorithm in different scenarios and the efficient inversion of the diagonally dominant matrix in precoding process.At the same time,in order to ensure that the system performance is not affected,a better precoding scheme is proposed,and finally,the software based millimeter wave 5G signal source is realized by combination of the dynamic link library and the user interface.The method verifies the feasibility and reliability of the precoding schemes proposed in this paper,and further improve the development of precoding technology in massive MIMO.The details are as follows:a)Theoretical analysis and verification of the precoding algorithm selection in different scenarios.The theoretical analysis of the three system models in massive MIMO technology was carried out,and the mathematical expressions of the system capacity under each model were derived.Based on the multi-user system,the performance of common linear precoding schemes was studied.The mathematical expressions corresponding to the precoding matrix and the received signal in different schemes were presented.Finally,the performance evaluation of the above precoding schemes was studied based on the bit error rate and system capacity.The theory and simulation analysis resulted in different features of each precoding algorithm in the application scenario.b)The study of the inversion of high-order matrices in linear precoding schemes.Using the corresponding mathematical analysis,the characteristics of the inverse matrix obtained in the precoding process were summarized.At the same time,the inverse matrix estimation algorithms were evaluated and compared with the conventional and relatively accurate inversion.However,these methods are too complicated in practical application with insufficient hardware resources.In order to solve the existing problems,this paper proposes a hybrid Neumann-Chebyshev precoding scheme with optimized performance,which combines the fast convergence of Neumann series expansion with the optimal search direction provided by Chebyshev iteration.By analyzing the convergence,applicability and computational complexity of the scheme,it is further verified that the hybrid scheme can perform faster convergence inversion with respect to other algorithms.c)The software implementation of the precoding schemes and the inversion algorithms.In order to apply the above methodology to a real millimeter wave 5G signal source prototype,the common linear precoding schemes and matrix inversion algorithms were implemented in software.The specific software functions are packed into dynamic link library with specifically designed function interfaces.On one hand,the parallel development of each functional module was realized,on the other hand,the coupling between the codes is reduced.It can resolve the problem that occurs during large scale program modification in terms of parameter change when combined with prototype engineering.Meanwhile,Qt Creator was used to design and develop the software user interface.The visual interface operation makes it more convenient for experimental data analysis and comparison,also this work moves towards the phase of functional test of the signal source prototype.This paper combines theory and simulation to obtain the advantages and disadvantages of existing precoding schemes under different application scenarios.The hybrid Neumann-Chebyshev precoding scheme was studied and analyzed.The scheme ensures the prerequisite of fast convergence and high computational accuracy with optimized performance.Finally,the software implementation of the corresponding precoding schemes and inverse matrix estimation algorithms were completed.The error analysis of the software calculation and the simulation data verify the practicability and reliability of the software.
Keywords/Search Tags:Massive MIMO, Precoding Algorithm, Matrix Inversion, Software Design
PDF Full Text Request
Related items