认知车联网中高效频谱探测与资源管理技术研究

Internet of Vehicles (IoV) is the most important component of intelligent transportation system. With the rapid growing of the road traffic density, the efficiency and reliability of data transmission in IoV are serious affected by the spectrum shortage. In this project, we integrate cognitive radio into IoV, and construct the CR-IoV. In the CR-IoV, the vehicles can sense the spectrum environment, and reconfigure the system to obtain new spectrum resource. However, for the uncertainties of the available spectrum resources and the fast moving of vehicles, there are three problems should be solved urgently in the CR-IoV, including: the cognitive system modeling is difficult, the spectrum detection method is inefficient, and the spectrum resource management technology is backward. In order to address these problems, in this project, according to analysis of the moving characteristics of vehicles and the different service demands of users, we construct the cognitive system model of the CR-IoV firstly. Then, we propose two efficient cooperative spectrum detection methods: vehicle-to-vehicle grouping method and database based vehicle-to-roadside infrastructure method. Finally, we develop two efficient multi-users spectrum resource management schemes: double-layer networks model based scheme and imperfect communication based scheme. In addition, for the convenience of engineering practice, we plan to design a set of low complexity algorithms for the proposed methods and schemes, by combining the intelligent optimization theory. The project provides the new perspectives to integrate the wireless communication，intelligent transport system technologies and intelligent optimization theory.

车联网是智能交通系统的主要实施载体。随着路面车辆密度的急剧增长，频谱资源短缺已严重影响车联网内数据传输的高效性和可靠性。本项目将认知无线电技术引入车联网，构建认知车联网。车辆对频谱环境进行认知，并重构系统以获得新的频谱资源。然而，可用频谱资源的不确定性和车辆快速移动的特征，使得认知车联网面临：系统认知模型精准建模难、频谱探测方法低效以及资源管理技术滞后等问题。针对上述三大问题，本项目首先对车辆的移动特征和用户差异化的服务需求进行分析，构建认知车联网的认知系统模型；然后，基于认知模型，依次设计路面车辆分组车与车、和基于数据库的车与路侧设施两种高效协作频谱探测方法；最后，分别提出在双层网络模型下、和“非理想”环境下两种多用户频谱资源管理策略。为了工程应用，本项目将结合智能优化算法，设计各项理论研究相应的低复杂度算法，探索通信技术、智能交通技术与智能优化理论交叉融合的新方式。

车联网是智能交通系统的关键实施载体。本项目对认知车联网展开研究，将“认知”扩展为对各类资源进行探测，针对车联网内各类资源进行有效地探测和高效管理。车辆的快速移动和可用资源的不确定性导致认知车联网存在：资源探测方法低效和资源管理技术滞后等问题。基于此，本项目以资源探测和协同管理为核心思想，结合5G无线通信、移动边缘计算、能源互联网等最新理论方法，对车联网内通信、计算、电力等资源进行有效探测并联合优化管理。具体研究工作包括：基于边缘计算的车联网资源探测与管理技术、不同场景下电动汽车电力资源调度技术、不可靠通信环境下车联网内通信传输干扰控制方法和能源互联网内电力资源共享策略。此外，为了满足应用需求，本项目结合智能优化算法，设计了各项所提策略的低复杂度算法，探索通信技术、智能交通技术与智能优化理论交叉融合的新方式。