【作者】 李靖；

【导师】 方俊；

【作者基本信息】 电子科技大学 ， 信息与通信工程， 2023， 硕士

【摘要】 随着无线通信技术的发展,大规模天线阵列的多输入多输出(Multiple-Input Multiple-Output,MIMO)技术作为一项高效的无线传输技术,能够充分利用空间复用资源,使无线通信系统获得更高的传输速率与频谱效率。然而,由于无线系统日益复杂,天线阵列的尺寸也在增大,如何在实际环境中部署尽可能多的天线阵元,已成为限制大规模MIMO技术应用的重要因素之一。此时研究密集化阵元的大规模天线阵列有望解决这一问题,一方面,制约密集化阵列性能的重要因素是阵列间的耦合效应;另一方面,阵元密集化导致的信号反射等非理想因素也会对系统性能造成损耗。有研究工作提出了一种阵列耦合可重构的天线架构,该架构对传统数字预编码技术提出了新的挑战,联合设计可控的耦合参数与数字预编码,实现无线通信系统性能优化是目前待解决的问题。本文针对密集化大规模天线阵列通信系统,在阵列耦合可重构的架构下,以多用户下行通信过程为研究场景,将耦合效应建模为可调阻抗的函数形式,分别设计可调阻抗与数字预编码提升系统性能。针对上述目标,以下行系统的频谱效率最大化为目标函数建立了优化问题。对于这一非凸优化问题的求解,采用了交替迭代的优化框架。利用加权最小均方误差(Weighted Minimum Mean Square Error,WMMSE)算法求解预编码方案;针对可调阻抗优化问题,在交替方向乘子法(Alternating Direction Method of Multipliers,ADMM)的框架下,构造多个具有局部凸结构的子优化问题进行迭代求解。最后,仿真验证了所提联合优化算法的性能优势,结果表明算法性能显著优于基线方案与对比算法。在此基础上,本文还考虑了非理想阵列情况,引入信号反射损耗这一非理想因素,将反射损耗表示为关于可调阻抗的函数形式。仍然以下行频谱效率最大化为优化目标,在交替迭代的优化框架下设计预编码与可调阻抗,一方面基于WMMSE优化算法设计了预编码方案;另一方面,将耦合效应优化与反射损耗抑制作为优化目标,提出了两种可调阻抗的优化算法。首先,在全空间中约束反射信号能量上界,利用ADMM构造子问题求解;随后,提出了一种两阶段优化算法,第一阶段,通过最小化反射损耗构造子搜索空间,第二阶段,将耦合效应的优化问题转化为二阶锥规划问题求解。仿真结果表明所提算法实现了对系统可达频谱效率的优化,有效地抑制了信号反射带来的能量损耗,相比基线方案取得了更优的性能增益。更多还原

【Abstract】 With the development of wireless communication systems,Massive Multiple-Input Multiple-Output(MIMO)technology,as an efficient wireless transmission technology in multi-antenna arrays,can fully utilize spatial multiplexing resources and enable wireless communication systems to achieve higher transmission rates and spectral efficiency.However,as the complexity of wireless systems increases,the size of antenna arrays also grows accordingly.To deploy as many antenna elements as possible in the actual environment has emerged as a significant bottleneck impeding the implementation of massive MIMO technology.Intensive research on large-scale antenna arrays with densely packed elements holds promising solutions to address this issue,On the one hand,a significant factor that constrains the performance of densified arrays is the coupling effect.on the other hand,non-ideal factors such as signal reflections caused by the densification of the arrays can also lead to losses in system performance.Some research work proposes a tunable antenna architecture with reconfigurable coupling to achieve coupling effect regulation,which poses new challenges to traditional digital precoding techniques.Now the joint design of controllable coupling parameters and digital precoding,with the goal of optimizing the performance of wireless communication systems,is a key issue at present.In this thesis,for dense large-scale antenna array communication system,under the architecture of array coupling reconfigurable,the coupling effect is modeled as a function of tunable load impedance in the scenario of downlink multi-user system.To improve the system performance,the tunable impedance and digital precoding schemes are designed jointly.For the above objective,an optimization problem is formulated with the objective of maximizing the spectral efficiency of the downlink system.An alternating iterative framework is implemented to solve this non-convex optimization problem.The Weighted Minimum Mean Square Error(WMMSE)algorithm is used to solve the precoding scheme; for the optimization problem of tunable impedance,multiple subproblems with locally convex structures are constructed for iterative solution in the framework of the Alternating Direction Method of Multipliers(ADMM).Finally,the numerical results on the performance of the proposed algorithms show that the proposed algorithm outperforms than the baseline scheme and the existing algorithms.On this basis,the non-ideal array scenario is taken into account and the non-ideal factor of signal reflection loss is incorporated by expressing it as a function of the tunable impedance.Firstly,the precoding scheme is designed based on WMMSE optimization algorithm.Subsequently,two optimization algorithms for tunable impedance are proposed with coupling effect optimization and reflection loss suppression as the objectives.In the first scheme,the upper bound of the reflected signal energy is constrained and the subproblem is constructed using ADMM to solve; then,a two-stage optimization algorithm is proposed,wherein the first stage involves the construction of the subspace by minimizing the reflection loss,while the second stage transforms the optimization problem of coupling effect into a second-order cone programming problem to slove.The simulation results show that the proposed algorithm achieves the optimization of the spectral efficiency and effectively suppresses the loss caused by the signal reflection,which achieves better performance gain compared with the baseline scheme.更多还原