毫米波MIMO系统信道估计与预编码技术研究

发布时间：2018-05-11 12:44

本文选题：毫米波 + MIMO　；参考：《电子科技大学》2017年硕士论文

【摘要】：未来的无线通信的目标是提供更高的传输速率,更大的系统容量,满足万物互联的通信需求。在传统的低频段频谱上实现这一目标并不现实,通信必须向高频段的毫米波扩展。毫米波的物理性质决定了其需要和MIMO技术联合应用,以提供良好的服务。在毫米波系统中使用大规模天线阵列,在提高性能的同时,也带来了经济成本和电路设计方面的问题。如果采用数字预编码方案,需要配置和天线数相等的射频链路,实现难度很大。采用混合架构可以解决这一难题。在射频域配置由移相器网络组成的模拟预编码器,在基带配置传统的数字预编码器,可以大幅减少系统需要的射频链路的数量。混合架构的毫米波MIMO系统在信道估计和预编码方面有很多值得研究的问题。第一,天线规模比较大,信道估计规模也同步上升,有必要研究新的信道估计方法。第二,信道估计准确度的提升也是一个目标。第三,模拟预编码器由移相器组成,移相器只能调节相位,不能调节幅值,这种物理限制因素决定了必须研究适用于混合架构系统的预编码方法。针对这些问题,本文开展了一些研究。本文首先研究了毫米波在散射空间中的传播特性,讨论了毫米波信道的时域稀疏性和角域稀疏性。基于信道的稀疏性,本文使用群簇模型对毫米波窄带信道进行了建模分析。建立了混合架构毫米波MIMO系统的数学表示模型。鉴于毫米波MIMO信道是稀疏的,本文对基于稀疏信号重构的信道估计方法开展了研究。首先,介绍了压缩感知和稀疏信号重构的相关知识,讨论了OMP和CoSaMP重构算法。基于这两种算法,本文设计了基于OMP和CoSaMP算法的信道估计方法,并进行了仿真,分析了两种方法的性能。在此基础上提出一种改进的稀疏信号重构算法,该算法与OMP的不同之处在于每次匹配时选择多个感知原子用于筛选。仿真结果表明,改进的信道估计方法的性能优于基于OMP的信道估计方法,且计算复杂度基本没有增加。最后,本文对适用于混合架构毫米波MIMO系统的预编码方法进行了研究。建立了混合预编码问题的数学模型并讨论了该问题的优化目标函数。在此基础上研究了两种基于交替优化算法的混合预编码矩阵计算方法。仿真结果表明,两种方法都可以实现良好的性能,且实现复杂度较低。
[Abstract]:The goal of future wireless communication is to provide higher transmission rate and larger system capacity to meet the communication needs of interconnection. It is not realistic to achieve this goal in the traditional low frequency spectrum, so the communication must be extended to the millimeter wave in the high frequency band. The physical properties of millimeter wave make it necessary to combine with MIMO technology to provide good service. The use of large-scale antenna arrays in millimeter-wave systems not only improves performance, but also brings economic cost and circuit design problems. If the digital precoding scheme is adopted, it is difficult to configure the RF link with the same number of antennas. A hybrid architecture can solve this problem. The configuration of analog precoder composed of phase shifter network in RF domain and traditional digital precoder in baseband can greatly reduce the number of RF links required by the system. There are many problems in channel estimation and precoding for hybrid millimeter-wave MIMO systems. First, the antenna size is large and the channel estimation scale is rising simultaneously. It is necessary to study a new channel estimation method. Second, the improvement of channel estimation accuracy is also a goal. Thirdly, the analog precoder is composed of phase shifters, which can only adjust the phase, but not the amplitude. This kind of physical limitation determines that a precoding method suitable for hybrid architecture systems must be studied. In view of these problems, this paper carried out some research. In this paper, the propagation characteristics of millimeter-wave in scattering space are studied, and the time-domain sparsity and angular sparseness of millimeter-wave channel are discussed. Based on the sparsity of the channel, the cluster model is used to model the millimeter wave narrowband channel. A mathematical representation model of a hybrid millimeter-wave MIMO system is established. In view of the sparse millimeter-wave MIMO channel, the channel estimation method based on sparse signal reconstruction is studied in this paper. Firstly, the knowledge of compressed sensing and sparse signal reconstruction is introduced, and OMP and CoSaMP reconstruction algorithms are discussed. Based on these two algorithms, the channel estimation methods based on OMP and CoSaMP algorithms are designed and simulated, and the performance of the two methods is analyzed. On this basis, an improved sparse signal reconstruction algorithm is proposed. The difference between this algorithm and OMP is that multiple perceptual atoms are selected for filtering each match. Simulation results show that the performance of the improved channel estimation method is better than that of the channel estimation method based on OMP, and the computational complexity is not increased. Finally, a precoding method for hybrid millimeter-wave MIMO systems is studied in this paper. The mathematical model of hybrid precoding problem is established and the optimization objective function of the problem is discussed. On this basis, two hybrid precoding matrix calculation methods based on alternating optimization algorithm are studied. The simulation results show that both methods can achieve good performance and low implementation complexity.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN919.3

【参考文献】