随着通信技术的演进和物联网产业的发展，各种室内定位技术不断涌现。其中，基于第五代移动网络通信（5G）的定位技术受到了广泛关注，这得益于其无需额外部署，大带宽，多天线等突出的优势。传统的基于蜂窝网络的定位技术是基于小区识别码的方式，其定位精度较低，无法应用在室内。在最新的5G网络协议中，规定了定位相关的上行和下行参考信号，其中上行的探测参考信号（SRS）具备良好的相关特性和幅度特性，有利于获得更准确的信道估计和定位。由于网络侧基站计算能力强，天线数目多，因此本文考虑利用SRS来进行5G定位系统的搭建和测角算法的设计。不过目前基于SRS信号的5G定位技术的研究大多基于仿真，尚无从实际系统中提取出物理层SRS符号进行定位的工作。另外，目前5G信号测角算法大多使用传统的窄带算法，无法应用到多载波宽带SRS信号的测角中，且大多算法尚未在商用设备上进行评估和验证。因此本文的工作主要旨在解决以上问题。本文提出了一种在实际商用设备上采集5G SRS物理层符号的方法，并基于5G标准流程和SRS配置参数生成发射端SRS，进行信道估计得到信道频域响应。本文根据硬件设备的特点建模了频域响应的测量模型，并针对到达角度（AOA）估计问题提出了信道频域响应的校正方法。实验验证，收发SRS信号符合标准，且提出的方法能有效校正天线间的信道响应相位差。本文提出了一种聚焦类MUSIC测角算法，来进行5G SRS信号的AOA估计。算法分为两个步骤，首先利用类MUSIC算法进行预估，再利用设计的聚焦算法将SRS信号各个子载波上的信息聚焦到参考频率上，从而充分利用宽带信息。实验结果表明，提出的测角算法在单径场景（微波暗室）可达到99%的情况下5°以内测角精度，在多径场景（室内实验室）可达到95%的情况下10°以内测角精度。本文搭建了基于5G商用设备的定位系统，包括终端侧的手机和网络侧的室内皮基站以及核心网。基于此系统，本文提出了5G SRS单基站定位方案并进行性能评估。在单基站定位方案中，利用SRS的角度估计以及接收信号强度得到的距离估计，在单径和多径距场景中定位中值误差为0.59m和1.51m。因此，本文提出的定位方案可在商用设备上部署，可以为未来的室内定位系统的部署和室内定位技术的研究提供产业化思路和算法基础。

With the evolution of communication technology and the development of the Internet of Things (IoT), various indoor positioning technologies are emerging. Among them, the positioning technology based on the fifth generation communication network (5G) has attracted wide attention due to its outstanding advantages such as low cost, large bandwidth and multiple antennas. The traditional positioning technology based on the cellular network is the cell ID, which can not be applied to indoor applications due to its low positioning accuracy. In the protocol of the 5G network, the uplink and downlink reference signals related to the positioning are specified. Due to the great correlation and amplitude characteristics of uplink sounding reference signal (SRS) and high capability as well as multiple antennas of base stations, the thesis considers the SRS to establish the 5G positioning system. Some essential and practical problems are usually ignored in current work. For example, the 5G communication systems will not record the received SRS symbols in the physical layer, which is essential for angle of arrival (AOA) algorithms. Furthermore, the current 5G AOA algorithms mostly are designed for narrow-band signals, which can not be applied to multi-carrier SRS and have not been verified on commercial equipment. So the thesis is mainly aimed at solving these problems. This thesis proposes a method to collect the physical layer symbols of 5G SRS on commercial equipment and generates the transmitted SRS based on the protocols and SRS configuration parameters. Furthermore, the thesis performs channel estimation to obtain the channel frequency response (CFR). In this thesis, the measurement model of the CFR is built according to the characteristics of the hardware equipment, and a calibration method for the CFR is proposed for the estimation of the AOA. Experimental results show that the generated and collected SRS are correct, and the proposed method can effectively calibrate the phase difference of the CFR.In this thesis, a focusing MUSIC-like algorithm is proposed to estimate AOA of the 5G SRS signal. The algorithm is divided into two steps. Firstly, the MUSIC-like algorithm is used for pre-estimation, and then each sub-carrier of the SRS is focused to the reference frequency by the designed focusing algorithm to exploit the wide-band information. Experimental results show that the proposed algorithm achieves 5 degrees in 99\% cases for the single-path scenario (microwave darkroom) and 10 degrees in 90\% cases for the multi-path scenario (indoor laboratory).This thesis establishes a 5G SRS positioning system based on commercial equipment including a mobile phone and base stations. The experiments are taken in a 10m\times10m\times3m typical indoor scenarios. Specifically, this thesis proposes a single base station positioning scheme, and the median positioning errors are 0.59m and 1.51m in single-path and multi-path scenarios by utilizing the AOA and the received signal strength of the SRS. Therefore, the positioning scheme proposed in this thesis can be deployed on commercial equipment, which can provide insights for the deployment of indoor positioning systems and the development of indoor positioning technologies.