摘要大规模MIMO系统能够提高频谱效率、能量效率和系统的稳定性,目前已经成为 5G无线通信领域最有潜力的研究方向之一。 本文重点研究了大规模 MIMO 系统的信道估计、压缩感知理论中稀疏重建算法在信道估计中的应用,以及毫米波大规模 MIMO 系统波束选择。波束选择在没有明显性能损失的情况下可以显著减少射频链路的数目。主要研究工作如下: (1)研究了基于导频的信道估计中的最小二乘(Least Squares,LS)、线性最小均方误差法(Minimum Mean-Square Error, MMSE)和奇异值分解线性最小均方误差法(Singular Value Decomposition- Minimum Mean-Square Error,SVD-LMMSE) ,并在OFDM信道进行仿真,结果显示性能从优到劣依次为:MMSE、SVD-LMMSE、LS,而复杂度由低到高依次为:LS、SVD-LMMSE、MMSE。 (2)调研了压缩感知理论中的稀疏重建算法:正交匹配追踪(Orthogonal Matching Pursuit,OMP)算法、正交最小二乘(Orthogonal Least Squares,OLS)算法、子空间追踪(Subspace Pursuit,SP)算法,在OFDM信道进行仿真,结果显示误码率和均方误差由低到高依次为:OLS、OMP 和SP,运算复杂度由低到高依次为:OMP、SP 和OLS。 (3)针对毫米波大规模 MIMO系统进行了研究,在考虑到多用户干扰的情况下,给出一种低复杂度的波束选择方案,来减少射频链路的数目,并在多径信道模型下仿真,结果显示此方案比传统波束选择方案有更高的能源效率,并且可以达到近似最优的和速率。74536
毕业论文关键词 大规模MIMO 系统,信道估计,稀疏重建算法,波束选择,最大和速率
Title Sparse Reconstruction Algorithm and Its Application in Massive MIMO System
Abstract Massive MIMO system can improve spectrum efficiency, energy efficiency and stability of the system。 It has become one of the most promising research directions in 5G wireless communication。 In this paper, Our important works include the following three aspects: the channel estimation in massive MIMO system, the application of sparse reconstruction algorithm in the theory of compressed sensing to wireless sparse channel estimation in massive MIMO, and the use of beam selection in millimeter wave massive MIMO system。 Beam selection can reduce the number of RF chain without obvious performance loss。 The main works are as follows: (1) We investigate pilot-based methods of the channel estimation which include least squares (LS), minimum mean square error (MMSE), singular-value-decomposition-linear MMSE (SVD-LMMSE)。 We simulate the three methods in OFDM system。 Simulation results show that the mean square error (MSE) and bit error rate (BER) performances of the three algorithms are in decreasing order: MMSE, SVD-LMMSE, and LS, and their complexities are in increasing order: LS、SVD-LMMSE、MMSE。 (2) We study the four channel estimators based on sparse reconstruction algorithms in compressing sensing theory: OMP(Orthogonal Matching Pursuit,OMP), OLS(Orthogonal Least Squares,OLS), SP(Subspace Pursuit,SP)。 Simulation results show that their performances of MSE and BER are in decreasing order: OLS, OMP, and SP, and their complexities are in increasing order: OMP, SP, and OLS。 (3) We investigate the beam selection of the millimeter massive MIMO system。 Taking into account the multi-user interference, a low-complexity beam selection scheme is presented to reduce the number of RF chains。 In multipath model, simulation results show that the low-complexity beam selection has higher energy efficiency than conventional approaches and it can achieve near-optimal sum rate。 稀疏重建算法及其在大规模MIMO系统中的应用:http://www.youerw.com/tongxin/lunwen_85173.html