传统依赖于电磁波电场强度信号的无线通信技术，面临着日益增长的用户传输速率需求与紧张有限的频谱资源的矛盾。作为无线通信的新维度，电磁波轨道角动量（OAM）为无线技术的革新提供了新方向，受到越来越广泛的关注与研究。本文围绕量子态OAM无线通信的关键问题——“OAM信号检测与量子态OAM涡旋电磁波的无线传输方法”展开相应研究工作，具体包括基于畸变图像识别的OAM检测方法，基于二维电子分选的OAM检测方法，以及量子态OAM的无线传输架构及其应用在隐蔽传输场景下的具体方法三个方面的工作。首先，针对涡旋微波量子接收时涡旋电子分布易发生畸变的问题，对非理想纵向磁场造成的非线性畸变进行了分析和建模，明确了非线性畸变会带来OAM模态间串扰，降低OAM模态纯度。因此，提出了一种基于卷积神经网络（CNN）的OAM模态识别方法，可有效对抗剩余磁场的影响，将OAM模态识别准确率从传统方法的71.88%提高到96.88%。其次，针对不同OAM模态涡旋电子的分选问题，提出了一种基于对数极坐标变换（LPCT）的二维涡旋电子分选方法。传统的基于LPCT的一维分选方法仅考虑了涡旋电子的螺旋相位特征，所提方法在其基础上引入与径向分布特征相关的径向相位调制，从而将涡旋电子分选到依赖于相位和径向分布的二维空间。不仅可以将模态间分选距离从一维分选方法的μm级增加到mm级，而且可以有效降低模态间串扰，相应分选准确率从80.92%增加到99.52%。其三，针对采用了涡旋微波量子的量子态OAM无线传输系统，提出了基于前述量子态OAM检测方法的OAM键控（OAM-SK）传输和OAM复用传输（OAM-DM）架构；进一步针对隐蔽传输场景，提出一种隐藏于广播信号之下的量子态OAM隐蔽传输方法，解决了统计态OAM隐蔽传输方法易通过相位梯度泄露信息的难题，实验验证了所提方法的有效性。
Traditional wireless communication technologies which rely on the electric field strength are confronted with the contradiction between the increasing transmission rate demand and the limited spectrum resources. As the new dimension of electromagnetic waves in wireless communications, Orbital Angular Momentum (OAM) provides a new insight into the innovation of wireless technologies, and attracts more and more attention.This dissertation focuses on the key issues of the quantum OAM wireless communication, such as the OAM recognition detection and the transmission method of vortex microwaves with quantum state OAM. Specifically, it includes three aspects, i.e., OAM identification detection method based on vortex electron distortion diffraction image identification, OAM detection identification method based on two-dimensional (2D) vortex electron sorting, the wireless transmission method of quantum state OAM and its application in covert transmission.Firstly, for the problem that vortex electrons are prone to be distorted when the vortex microwave photons are received, the nonlinear distortion caused by the nonideal longitudinal magnetic field is analyzed and modeled. It is clear that nonlinear distortion will bring crosstalk between different OAM modes and reduce the OAM purity. An OAM recognition method based on Convolutional Neural Network (CNN) is proposed, which can effectively resist the influence of nonideal magnetic field.recognizes OAM modes by using vortex electron diffraction images. Even for vortex electrons with nonlinear distortion caused by magnetic field, this This method can improve the OAM recognition accuracy from 71.88% to 9496.1288%.Secondly, for the separation of vortex electrons in different OAM modes, a two-dimensional (2D) vortex electrons sorting method based on Log-Polar Coordinate Transformation (LPCT) is proposed. The traditional one-dimensional (1D) LPCT-based sorting method only considers the helical phase characteristics of the vortex electrons. The proposed method introduces the radial phase modulation related to the radial distribution characteristics, and the vortex electrons can be sorted into a two-dimensional space dependent on the phase and radial distribution. The sorting distance between adjacent modes can be increased from μm level of the 1D method to mm level, and the crosstalk between OAM modes can be effectively reduced. The corresponding sorting accuracy can be increased from 80.92% to 99.52%.Thirdly, for the quantum state OAM wireless transmission system using vortex microwave photons, an OAM Shift Keying (OAM-SK) transmission and an OAM Division Multiplexing (OAM-DM) transmission architectures based on the aforementioned quantum state OAM detection technologies are proposed. Furthermore, for covert transmission scenarios, considering that the statistical state OAM covert transmission is easy to leak information through phase gradient, a quantum state OAM covert transmission method hidden under broadcast signal is proposed, which can solve the problem that the statistical state OAM covert transmission is easy to leak information through phase gradientobtain the improvement of transmission rate and security simultaneously. The effectiveness of the proposed method is verified by experiments.
