分布式MIMO天线系统的时延估计算法研究

发布时间：2018-08-09 11:43

【摘要】：分布式MIMO(distributed multiple-input multiple-output,D-MIMO)是目前探索第五代移动通信(the fifth generation mobile communication,5G)中的关键技术之一,具有高容量、高频谱利用率、低功耗、更好的覆盖、开放式的结构等优点。时延估计一直是通信等领域研究的热点之一,直接决定着D-MIMO天线系统整体性能的好坏。信号检测、信道均衡、解调等信号处理操作都依赖于与时延估计密切相关的定时同步和信道估计。信道环境通常是多径的,目前主流的基于前导的帧同步方法估计的是第一径为主径且功率较大情况下的时延。而信道估计中冲激响应的位置即为估计的多径相对时延。因此,本文将从多径时延估计和帧同步两个方面探究分布式MIMO天线系统的时延估计。分布式压缩感知(Distributed Compressed Sensing,DCS)是压缩感知的扩展,针对多个信号的联合稀疏性差异,主要有三种联合稀疏模型及其重构算法,能够利用较少的观测数实现多个信号的联合重构估计,已经被很好地应用在通信系统多个稀疏信道的联合估计中。ADMM(Alternating Direction Method of Multipliers)是一种很受欢迎的分布式并行计算的优化求解方法,能够在每次迭代都达到近似最优解且计算速度快。对于分布式MIMO天线系统中,多径时延估计的精度不高、计算复杂度较大、分辨力不够等问题,结合多径信道的相关性大小和稀疏性,本文提出了基于ADMM的DCS多径时延估计方法,包括估计模型的建立、目标函数的确定以及算法求解等步骤。通过仿真实验表明所提的方法不仅具有较好的时延估计性能,而且提高了分辨力,节省了测量数目从而减小了计算复杂度。针对传统分布式MIMO天线系统估计第一径到达时延的帧同步算法中,一般采用遍历搜索的方法确定正确同步索引点,其存在计算复杂度和计算量较大的问题。而黄金分割法是求取一维极小化问题的近似最优策略,且具有计算复杂度低的特点。本文从同步度量函数曲线的几何特征出发,针对定时同步估计经典算法遍历搜索计算量大的问题,提出一种新颖的基于黄金分割法优化的定时同步算法,以其中的帧同步及在分布式MIMO系统的应用为例给出了算法流程和复杂度分析。通过仿真实验,验证了所提方法在大幅降低复杂度的情况下,仍然保持了与传统算法相当的性能。
[Abstract]:Distributed multiple-output D-MIMO (D-MIMO) is one of the key technologies in exploring the (the fifth generation mobile communication 5G of the fifth generation mobile communication. It has the advantages of high capacity, high spectral efficiency, low power consumption, better coverage and open architecture. Time delay estimation is one of the hotspots in the field of communication, which directly determines the overall performance of D-MIMO antenna system. Signal processing operations such as signal detection, channel equalization and demodulation depend on timing synchronization and channel estimation, which are closely related to time delay estimation. The channel environment is usually multipath, and the current leading based frame synchronization method is used to estimate the time delay in the case of the first path being the main path and the power being high. The position of impulse response in channel estimation is the estimated multipath relative delay. Therefore, this paper will explore the time delay estimation of distributed MIMO antenna systems from two aspects: multipath delay estimation and frame synchronization. Distributed compressed sensing (Distributed Compressed) is an extension of compression sensing. There are mainly three joint sparse models and their reconstruction algorithms for the difference of joint sparsity of multiple signals, which can realize the joint reconstruction estimation of multiple signals with fewer observations. ADMM (Alternating Direction Method of Multipliers) has been well applied to the joint estimation of multiple sparse channels in communication systems. ADMM (Alternating Direction Method of Multipliers) is a popular optimization method for distributed parallel computing, which can achieve approximate optimal solution in each iteration and high computation speed. For distributed MIMO antenna systems, the multipath delay estimation is not accurate, the computational complexity is large, the resolution is not enough, and so on. Considering the correlation and sparsity of multipath channel, a DCS multipath time delay estimation method based on ADMM is proposed in this paper. It includes the establishment of the estimation model, the determination of the objective function and the calculation of the algorithm. The simulation results show that the proposed method not only has better time delay estimation performance, but also improves the resolution, saves the number of measurements and reduces the computational complexity. In the traditional frame synchronization algorithm for estimating the first path arrival delay in distributed MIMO antenna systems, the traversal search method is generally used to determine the correct synchronization index points, which has the problem of large computational complexity and computational complexity. The golden section method is an approximate optimal strategy for solving the one-dimensional minimization problem and has the characteristics of low computational complexity. In this paper, a novel timing synchronization algorithm based on golden section method is proposed to solve the problem that the classical algorithm of timing synchronization estimation has a large amount of traversal search and computation based on the geometric characteristics of the synchronization metric curve. Taking frame synchronization and its application in distributed MIMO system as an example, the algorithm flow and complexity analysis are given. The simulation results show that the proposed method still keeps the same performance as the traditional algorithm under the condition of greatly reducing the complexity.
【学位授予单位】：湘潭大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN919.3

【参考文献】