摘要为了满足人们对于无线通信速率日益增长的需求，学者们提出了大规模MIMO技术。通过在基站配置大规模数量的天线阵列，大规模 MIMO 系统的能效性和频谱效率得到了显著提升。同时，由于采用大规模天线阵列后，大规模 MIMO 系统中各用户的信道矩阵趋于正交，用户间的相互干扰也得到了良好的抑制。目前，大规模 MIMO 技术的主要瓶颈问题之一是高复杂度。本文主要针对大规模 MIMO 关键技术波束成形的复杂度降低和性能分析展开了研究，主要工作如下： 1） 、研究了小规模 MIMO 系统中代表性预编码算法：块对角化算法 (Block Diagonalization, BD) ，匹配滤波算法，迫零算法 (Zero forcing, ZF) ，最小均方误差算法，最大化信泄噪比算法和奇异值分解算法 (Singular Value Decomposition, SVD) 。仿真结果表明：在和速率性能方面，SVD 性能最优，BD和 ZF 居中，MF性能最差；复杂度方面，MF复杂度最低，其它算法复杂度基本相同。 2） 、针对大规模 MIMO 系统中 ZF 预编码算法求伪逆复杂度高的问题，给出了三种低复杂度迭代算法：Jacobi、Gauss-Seidel（GS）和 SOR。仿真结果表明：在大规模 MIMO中，随着天线数量的增加，三种算法复杂度均明显低于传统 ZF 算法，性能趋近于 ZF 算法；三种算法中，GS 和 SOR算法的收敛快，仅需 4 次迭代就可达 ZF算法和速率性能。 41656
毕业论文关键词：大规模 MIMO，波束成形，迫零，复杂度，系统和速率
Title Research in Key Technology of Massive MIMO System
Abstract To satisfy people’s command for higher wireless transmission rate, massive multiple-input multiple-output (Massive-MIMO) system has been proposed by researchers. Due to a large-scale number of antennas, massive MIMO achieves far higher energy efficiency and the spectrum efficiency than conventional small-scale MIMO. Meanwhile, as the number of antennas of the system approaches infinity, every subchannel tends to be orthogonal with each other, which greatly suppresses the interference between users. However, there are still a few bottleneck problems in massive MIMO technology, one of which is its high complexity in bemaforming. This paper investigates how to reduce the complexity of zero-forcing (ZF) beamforming in massive MIMO system, respectively. This paper is organized as follows: 1) We make an investigation on representative precoders: block diagonalization, matched filtering (MF), zero forcing, minimum mean square error, maximum signal-to-leakage-and-noise ratio, and singular value decomposition (SVD). Our simulation and analysis find: the SVD is the best preccoder in sum-rate, the MF performs worst in sum-rate while its complexity is the lowest. 2) To reduce the high complexity of zero-forcing (ZF) in massive MIMO system, three low-complexity iteration algorithms are presented: Jacobi, Gauss-Seidel (GS) and Successive Over-relaxation (SOR). It follows from our simulation and analysis, as the number of antennas increases, the three procedures show a significant reduction in complexity compared to the matrix-inverse-based ZF but also achieves the same sum-rate performance as the latter. Among the three algorithms, GS and SOR show more rapid convergence rate compared to Jacobi, and only require four iterations to achieve the same sum-rate performance as ZF.
Keywords Massive MIMO, Beamforming, Zero Forcing, Complexity, Sum-rate
目次
1绪论1
1.1研究背景1
1.2大规模MIMO技术简介2
1.3波束成形技术3
1.4国内外研究现状4
1.5论文主要结构5
2线性预编码算法6
2.1引言6
2.2系统模型6
2.3经典线性预编码算法9 大规模MIMO网络关键技术研究:http://www.youerw.com/tongxin/lunwen_41775.html