【作者】 张春红；

【导师】 禹永植； 李志刚；

【作者基本信息】 哈尔滨工程大学 ， 电子与通信工程（专业学位）， 2022， 硕士

【摘要】 随着通信和数据业务的蓬勃发展,各类新型无线业务极大的优化了人们的生活品质。为了满足密集的用户需求,正在发展的无线通信系统提高了移动系统下的能量消耗,但更广的网络覆盖范围和功耗更低的无线通信系统也给无线通信系统带来了更多的挑战。2021年我国已初步建成了全球最大规模的第五代移动通信系统移动网络(5th Generation mobile communication systems,5G),标志着中国5G正式商用的时代已经正式开启。而被称作5G的关键技术之一的大规模多输入多输出(Multiple-Input and Multiple-Output,MIMO)利用将基站处配备数十乃至数百根天线的方法有效的实现了提升大规模MIMO系统的整体频谱效率和系统总能量效率。本文将针对多小区中非完美信道状态信息(Channel State Information,CSI)情况下的大规模MIMO系统下行链路中的频谱效率与能量效率的优化进行深入探讨。综上所述,针对膨胀的用户需求与业务要求和绿色能源的挑战,大规模MIMO系统中对于频谱效率和能量效率的优化问题已经达到了刻不容缓的地步,并且具有非常现实的研究意义。本论文所做的主要工作如下:1.针对多小区场景建立非完美CSI情况下的大规模MIMO系统的系统模型来研究分析下行链路的数据传输过程,提出一种有效提高系统频谱效率的实现方法,该方法包含下行链路信道估计、信号接收技术以及功率分配算法,共同考虑三者优化,最终得到可以实现频谱效率最大化的方法实现。首先利用经典信道估计最小均方误差算法(The Minimum Mean Square Error,MMSE)引出一种适用于多小区环境的信道估计算法MCMMSE(Multicell MMSE),MC-MMSE算法可以将观测值与信道估计矩阵得到对角矩阵的方式来使信号间相关性运算更精确,从而提高大规模MIMO的系统性能。然后采用MRT、ZF和MMSE信号接收技术进行下行链路的公式推导,建立最大化系统总频谱效率的优化方程;根据两种不同的功率分配方式来进行频谱效率的优化比较,仿真分析得到MC-MMSE信道估计技术比MMSE信道估计技术有更大的频谱效率优化空间,并且在三种信号接收技术下运用MC-MMSE信道估计技术都可以获得更好的频谱效率,最后得出结论运用MC-MMSE估计算法、等功率分配算法并搭配MRT信号接收技术的这一种最大化频谱效率的实现方法可以发挥出系统更好的性能,提高整体系统的频谱效率。2.利用随机几何方法建立了大规模MIMO系统的下行链路的系统模型,设计了一种能实现能量效率最大化的多小区大规模MIMO系统下行链路的实现方法,在系统模型中引入实际通信系统中导频复用系数、下行传输的固定比例、硬件损伤程度等参数,通过数学公式推导出包含基站的天线数、用户数、小区内基站密度、导频复用系数以及功率控制因子等参量的下行链路的信干噪比表达式从而得到频谱效率。然后,将能量效率设定为区域性质,考虑区域内发射功率、硬件损耗、信号处理等各种情况下能耗的功耗模型,由此得到能效方程。最后通过交替迭代的优化算法进行能效优化并得到在最大化能效时的各个参数的数值。经仿真验证得到,本文所提下行链路实现方法具有良好的收敛性和可行性,可以实现能效优化,仿真得到的基站天线数与用户数符合大规模MIMO系统的天线设置,且与其他多小区的下行链路能效相比,在能量效率方面也具有明显优势,数值方面提升了12.2%左右,并且在环境发生改变时本文所提方法具有更强的鲁棒性,在实际情况中也具有代表性。更多还原

【Abstract】 With the vigorous development of communication and data services,various new wireless services have greatly optimized people’s quality of life.In order to meet the needs of intensive users,the wireless communication system under development has increased the energy consumption of the mobile system,but the wider network coverage and lower power consumption wireless communication system also bring more challenges to the wireless communication system.In 2021,my country has initially built the world’s largest mobile network(5th Generation mobile communication systems,5G),marking the official start of China’s 5G commercial era.The massive multiple-input and multiple-output(MIMO),known as one of the key technologies of 5G,can effectively improve the massive MIMO system by equipping dozens or even hundreds of antennas at the base station.The overall spectrum efficiency and the total energy efficiency of the system.This article will conduct an in-depth discussion on the optimization of spectrum efficiency and energy efficiency in the downlink of a massive MIMO system in the case of imperfect channel state information(CSI)in multiple cells.In summary,in view of the expanding user needs and business requirements and the challenges of green energy,the optimization of spectrum efficiency and energy efficiency in massive MIMO systems has reached the point where it is urgent and has very practical research significance.The main work done in this paper is as follows:1.Establish a system model of a massive MIMO system under imperfect CSI for multicell scenarios to study and analyze the data transmission process of the downlink,based on the classic channel estimation minimum mean square error algorithm(The Minimum Mean Square Error,MMSE)A channel estimation algorithm MC-MMSE(multicell MMSE)suitable for a multi-cell environment.The MC-MMSE channel estimation algorithm can obtain a diagonal matrix from the observation value and the channel estimation matrix to make the correlation calculation between signals more accurate,thereby Improve the system performance of massive MIMO,derive and analyze MC-MMSE to get the estimated value and estimation error.Then three precoding technologies of MRT,ZF and MMSE are used to derive the formula of the downlink to establish the maximum total spectrum efficiency of the system.According to the two different power allocation methods,the spectrum efficiency is optimized and compared,and the MC-MMSE is obtained by simulation analysis.Channel estimation technology has a larger spectrum efficiency optimization space than MMSE channel estimation technology,and the use of MC-MMSE channel estimation technology under the three precoding technologies can obtain better spectrum efficiency.Finally,it is concluded that the water injection power allocation algorithm is used.And with MRT precoding technology,it can give play to the better performance of the system.2.The system model of the downlink of the massive MIMO system is established by using the random geometry method,and a method to realize the downlink of the multi-cell massive MIMO system that can maximize the energy efficiency is designed,and reality is introduced into the system model.In the communication system,parameters such as the pilot reuse factor,the fixed ratio of downlink transmission,the degree of hardware damage,etc.,are derived through mathematical formulas,including the number of base station antennas,the number of users,the base station density in the cell,the pilot reuse coefficient,and the power control factor,etc.Parametric expression of the signal-to-interference and noise ratio of the downlink to obtain the spectral efficiency.Then,the energy efficiency is set as the regional nature,and the power consumption model of the energy consumption in various situations such as the transmission power,hardware loss,and signal processing in the region is considered,thereby obtaining the energy efficiency equation.Finally,the EE optimization is carried out by the alternating iterative optimization algorithm and the numerical values of the various parameters when maximizing EE are obtained.After simulation verification,the downlink implementation method proposed in this paper has good convergence and feasibility,energy efficiency optimization can be realized,the number of base station antennas and the number of users obtained by simulation are in line with the antenna settings of large-scale MIMO systems,and compared with the energy efficiency of other multi-cell downlinks,it also has obvious advantages in energy efficiency,and the numerical value is increased by about 12.2%,and the method proposed in this paper has stronger robustness when the environment changes,and it is also representative in actual situations.更多还原