Research On RF Impairments In Massive MIMO System

In 5th Generation network,the millimeter wave band can provide a larger bandwidth for the network to meet the demand of 5G network in transmission capacity.However,due to the severe path loss in the millimeter wave frequency band and the response characteristics of atmospheric,rainwater,shadow,etc.,the transmission performance is limited.So the Massive Multiple Input Multiple Output system is used to carry out beamforming.The beamforming technology has became an important technology to improve the spatial diversity and multiplexing gain.Due to the current level of radio frequency device process,the increase in the number of radio frequency links due to the increase in the number of antennas will bring various imaginativeness to the transmission signal,such as:phase noise,I/Q unbalanced,and non-linear distortion of amplifiers.This Thesis uses millimeter-wave Massive MIMO communication as a research topic.It focuses on the effects of the performance of the millimeter-wave Massive MIMO system due to the phase noise which brought by the radio frequency chain,and uses two phase noise suppression methods to compensate it.This thesis uses the OFDM modulation method,considering the two different effects of phase noise on the OFDM symbols,two different methods are used to compensate them.First,uses a new interpolation method which is improved by the classic IPC method to compensate the CPE.Second,uses the variational Bayesian inference and the principal component analysis to compensate the ICI.The main research contents and results in this thesis are as follows:(1).Analysis of several antenna architectures used in the millimeter-wave Massive MIMO system,and analyzing the impact of different antenna architecture on the overall system.Subsequently,based on the basic model of phase noise,a system model of phase noise was established in the millimeter-wave Massive MIMO system which used the OFDM modulation.Through the theoretical derivation and simulation,research the affection of the phase noise to Massive MIMO system.(2).Based on the classic IPC algorithm,an improved interpolation method is proposed.This method first compares the estimated channel difference between the derivative position in the adjacent OFDM symbols,and then uses the interpolation algorithm to compensate the data carriers which is located in the same symbol different pilot carriers.Finally achieve the purpose of inhibiting the entire transmission data CPE.(3).Based on the variational Bayesian inference and the principal component analysis,uses two steps of rough compensation and precise compensation to compensate the phase noise.In the rough compensation stage,this algorithm uses the pilot signal to estimate the phase noise component.In the precise compensation,the algorithm first uses the principal component analysis to reduce the dimension of phase noise in frequency domain.And then uses the variational Bayesian inference to get the posterior mean and posterior variance to inhibition the ICI.Finally,the above two methods are simulated on the simulation platform,and the system performance is compared between the full digital architecture and the hybrid architecture.Experiments show that the two methods proposed in this thesis can improve system performance to varying degrees compared to some traditional methods.