【作者】 李敏；

【导师】 景小荣；

【作者基本信息】 重庆邮电大学 ， 电子与通信工程（专业学位）， 2017， 硕士

【摘要】 大规模多输入多输出(Multi-Input Multi-Output,MIMO)技术通过在基站侧部署大量天线单元,使得系统频谱效率、能量效率得到大幅度提升,因此,大规模MIMO技术作为下一代移动通信系统的候选关键技术而广受关注。然而要充分发挥大规模MIMO技术的性能优势,必须采取有效的措施去解决系统中的小区间干扰(Inter-Cell Interference,ICI)和小区内用户间干扰(Inter-User Interference,IUI)问题,而预编码作为一种信号预处理技术,能够有效地消除干扰,同时减少接收机设计的复杂度,因此被广泛研究。在基于时分双工(Time Division Duplexing,TDD)工作模式的大规模MIMO系统中,不但存在上述干扰问题,同时不可避免地还存在导频污染问题,如何消除因导频污染问题而引发的干扰,是本论文需要解决的其中一个问题。同时,在基于频分双工(Frequency Division Duplexing,FDD)工作模式的大规模MIMO系统中,用户通常需要通过反馈链路将信道状态信息(Channel State Information,CSI)反馈给基站,因而系统中存在较高的反馈开销问题,如何在减小反馈开销的基础上,解决系统内存在的ICI和小区内IUI,是本论文所要研究的另外一个问题。基于上述问题,本论文对大规模MIMO系统中的双层预编码设计方案展开研究。具体内容包括:1.结合导频污染消除的双层预编码方案设计在TDD工作模式,为了有效地消除系统中小区内IUI、ICI以及导频污染问题引发的干扰,本论文给出一种导频污染条件下的双层预编码方案。该方案首先基于正则化(Regularized Zero Forcing,RZF)迫零技术设计外层预编码矩阵来消除小区内IUI,进而将实际信道和外层预编码矩阵作为等效信道,基于最小化均方误差(Minimum Mean Square Error,MMSE)准则设计内层预编码矩阵以消除ICI以及导频污染问题引发的干扰。数值仿真结果显示,该双层预编码方案不但有效地消除了系统内的干扰,同时使得系统和速率得到显著提升。2.结合信道协方差矩阵的双层预编码方案设计在FDD工作模式,为了消除小区内IUI,减小系统反馈开销,本论文给出了两种双层预编码设计方案。其中一种基于梯度上升法来实现双层预编码方案设计,该方案的外层预编码矩阵根据各组用户的信道协方差矩阵并运用梯度上升法进行设计,以消除各组间的IUI,进而将实际信道与外层预编码矩阵作为等效信道,以实现信道降维,最后基站基于等效信道并运用迫零(Zero Forcing,ZF)技术消除组内的IUI;因此,该方案一定程度上降低了系统反馈开销。另外一种基于泰勒级数展开来实现双层预编码方案设计,即基站首先根据各组用户信道协方差矩阵并运用泰勒级数展开进行外层预编码矩阵的设计来消除各组间的IUI,接着将实际信道和外层预编码矩阵看成等效信道,实现信道降维,最后基站根据等效信道并运用ZF技术进行内层预编码矩阵的设计,消除组内的IUI。数值仿真结果表明,基于泰勒级数展开的预编码方案优于基于梯度上升法的预编码方案。更多还原

【Abstract】 Massive multi-input multi-output(MIMO)technology can greatly improve the spectrum efficiency and energy efficiency of system by deploying a large number of antenna elements at the base station.Therefore massive MIMO technology is regarded as a key alternative technology of the candidate of the next generation mobile communication system and attracts tremendous attention.However,in order to take full of advantage of massive MIMO technology,the inter-cell interference(ICI)and inter-user interference(IUI)must be completely eliminated.As a kind of signal preprocessing technology,precoding is widely studied that can effectively eliminate the interference and reduce the complexity of the receiver designs.On one hand,the interference problem and pilot contamination problems coexist in massive MIMO systems for time division duplexing(TDD)mode.Hence how to eliminate the interference and pilot contamination is taken as one of the problem to be studied in this thesis.On the other hand,the user usually needs to feed the channel state information(CSI)back to the base station through the feedback link resulting to a higher feedback overhead problem for the massive MIMO system for frequency division duplexing(FDD)mode.Hence how to reduce the feedback overhead with the reduced interferences of inter-cell and IUI in intra-cell of the system is taken as another problem to be studied in this thesis.In the point of these,this thesis focuses on the research on the dual-layer precoding design scheme in massive MIMO system.The corresponding contents can be outlined as follow.Dual-layer precoding scheme design combined with pilot contamination elimination.In order to effectively eliminate IUI,ICI and pilot contamination in the system,a dual-layer precoding scheme with pilot contamination elimination based on TDD mode is presented.Firstly,the scheme designs the outer precoding matrix based on regularized zero-forcing(RZF)technology to eliminate the IUI.Then based on minimization of mean square error criterion(MMSE),the actual channel and the outer precoding matrix is regarded as the equivalent channel to designs the inner precoding matrix in order to eliminate ICI and the interference caused by pilot contamination.The numerical simulation results show that the proposed scheme not only effectively eliminates the interference in the system,but also makes the sum rate of the system be improved significantly.Dual-layer precoding scheme design combined with channel covariance matrix.In order to eliminate the IUI and reduce the feedback overhead,two kinds of dual-layer precoding design schemes are given for FDD mode.The first scheme is based on the gradient ascent algorithm to achieve dual-layer precoding design.According to the channel covariance matrix of each group,the outer precoding matrix of this scheme is designed by utilizing gradient ascent algorithm to eliminate the IUI between the groups,and then the actual channel and the outer precoding matrix is taken as the equivalent channel to further reduce the channel matrix dimension.Finally,based on the equivalent channel,the zero-forcing(ZF)technique is utilized to eliminate the IUI.Thus,this scheme reduces the system feedback overhead to a certain extent.The other scheme design the dual-tier precoding scheme based on the taylor series expansion.In this scheme,according to each group of channel covariance matrices,the IUI between the groups is eliminated by the taylor series expansion,then the actual channel and the outer precoding matrix is regarded as the equivalent channel to further decrease the channel matrix dimension.Finally,the inner precoding matrix is designed to eliminate IUI based on the ZF criterion.The numerical simulation results show that the performance of the precoding scheme based on the taylor series expansion is better than the precoding scheme based on the gradient ascent algorithm.更多还原