毫米波车联网通信信道测量与建模研究

Internet of vehicles (IoV) is the information exchange network between vehicles and the outside world. It includes vehicle-to-vehicle (V2V), vehicle-to-infrastructure (V2I), vehicle-to-pedestrian (V2P), etc. and is called vehicle-to-everything (V2X) in general. Conventional vehicular communication technologies at sub-6 GHz band cannot meet the ever-growing requirements of Gbps level data transmission rate and ms level latency. Millimeter wave (mmWave) communication can efficiently address this challenge due to its large bandwidth and by the use of short packet transmission. As an application in 5G scenario of ultra-reliable low-latency communication (uRLLC) and a new paradigm in vertical industry, mmWave vehicular communication will largely promote the integration developments of 5G and vehicular communication network. However, current mmWave vehicular communication network lacks the verification of channel measurement results in real environments. Meanwhile, the channel models have not fully taken channel characteristics into consideration, such as the high path loss, weak diffraction ability, low antenna heights, transmitter (Tx) and receiver (Rx) movements, Doppler shift, and time-varying non-stationarity. To solve the problems, this project will conduct various mmWave vehicular channel measurement campaigns at multiple frequency bands and multiple scenarios. The channel propagation characteristics of mmWave V2V, V2I, and V2P channels will be analyzed thoroughly. In addition, a geometry based stochastic model (GBSM) which can accurately describe the channel characteristics will be proposed. The work can provide experimental and theoretical foundations for the design and evaluation of mmWave vehicular communication network, as well as support the fast developments of 5G and beyond 5G, future intelligent vehicles, automated driving, and intelligent transportation system (ITS) in China.

车联网是车辆与外界交换信息的网络，包括车与车(V2V)、车与路面基础设施(V2I)、车与行人(V2P)等，统称为V2X。传统的6 GHz以下车联网通信技术无法满足日益增长的Gbps级数据传输速率与毫秒级时延，而毫米波巨大的带宽与短包传输可以有效解决这一挑战。毫米波车联网通信作为5G移动通信超可靠低时延应用场景在垂直行业的新型范式，将极大促进5G车联网的融合发展。但是，目前的毫米波车联网通信系统缺乏真实环境信道测量结果的验证，信道模型也未充分考虑路径损耗大、绕射能力弱、天线高度低、收发端移动、多普勒频移、时变非平稳等信道特性。本项目针对这一难题，拟开展广泛的多频段多场景毫米波车联网通信信道测量实验，深入分析信道的传播特性，并提出准确描述信道特性的几何随机信道模型，为毫米波车联网通信系统提供设计与评估的实验与理论基础，支撑我国5G及5G之后、未来智能汽车、自动驾驶、智能交通运输系统的快速发展。

本项目针对毫米波车联网通信信道测量、特性分析与信道建模开展研究。取得的代表性成果如下：1）开展了28/32/39 GHz毫米波多频段多场景信道测量实验，研究了路径损耗、人体/车辆阻挡、时域非平稳性、空间一致性等信道特性；2）提出了通用毫米波/太赫兹多天线时变非平稳信道模型，研究了空时频相关函数等信道特性；3）提出了基于机器学习的预测信道模型，通过实测与仿真数据进行验证，研究了频段、场景配置等参数与信道特性之间的关系。项目发表期刊论文8篇，会议论文2篇，培养博士生5名，相关研究揭示了毫米波车联网通信信道的传播特性，提出了准确反映信道特性的毫米波车联网通信信道模型，为毫米波车联网通信系统提供设计与评估提供了实验与理论支撑。