基于车辆社交属性的车联网置信数据分发技术

Internet of vehicle (IoV) is born with the social attribute, which provides new solutions for the issue of data dissemination in the complex traffic system. The area, utilizing the social nature to solve the issues of IoV, is becoming a hot research direction. This proposal aims at improving the efficiency of the data dissemination technology in IoV. To address the important but rarely involved issue in trust-based data dissemination technologies, the proposal adopts the following steps. (1) Within the limitation of the network resource, we design the method with low overhead for the neighbor discovery, which achieves the optimal tradeoff between the discovery accuracy and the overhead, and provides accurate neighbor information for the data dissemination technology. (2) Utilizing the multidimensional vehicle social attributes, we construct the trust model depending on the Dempster-Shafer theory of evidence (DST), which solves the issue of trust and build the basis for estimating the belief degree in the data dissemination technology. (3) For the complex traffic system, we propose the trust-oriented data dissemination method based on the greedy opportunity forwarding algorithm and carry-and-forwarding algorithm, which considers the impacts of the realistic transportation scenarios in IoV and finally makes the trust-oriented data dissemination come true. Since IoV is one crucial component of the Intelligent Transportation Systems (ITS), the research achievements of the project will be the reliable theoretical basis and technical solutions for the implementation of IoV, as well as the powerful supporting for developing a new intelligent transportation system.

车联网表现出高度的社交属性，为解决复杂交通系统中的数据分发问题提供了新思路，逐渐成为车联网中的热门研究方向。本申请课题以高效的车联网数据分发技术为研究目标，针对网络中至关重要、却仍未解决的置信分发问题，通过：（1）突破有限网络资源限制，设计低开销的邻节点发现方法，实现邻节点发现精度与开销随网络动态的优化自适应，为置信数据分发提供精确邻节点信息保障；（2）利用多维车辆社交特性，建立基于Dempster-Shafer证据融合理论的动态置信评估模型，解决节点置信问题，为数据分发提供置信依据；（3）面向复杂交通系统，提出基于贪婪机会转发和存储转发的置信数据分发方法，解决车联网三维场景影响和间歇性连接问题，最终实现车联网中的高效置信数据分发技术。车联网是智能交通的重要组成部分，本申请课题的研究能为将来车联网落地计划提供可靠的理论依据和技术方案，为探索新型智能交通系统提供重要支撑。

人工智能、区块链、最优化理论等技术的发展推动着车联网技术的不断前进。经过三年的持续投入，项目组面向复杂的交通系统，从理论建模入手，经由随机分析、仿真验证，面向高效的车联网数据分发技术，在单车感知、车辆置信和多车通信各个击破：首先研究了车联网中无线传输环境、车辆和业务类型对网络资源的占用及网络协议性能的影响，并进一步提出了移动性和信道自适应的邻节点发现方法，为车辆单车感知提供了可靠技术保障；其次，对车联网领域的置信技术进行了全面系统的综述，在此基础上研究了区块链技术在车联网置信领域的应用，并提出了基于信任值更新的车联网信任机制，辅助车联网通信安全；最后，从网络架构、转发策略两方面出发，研究了提升车联网数据传输可靠性的方法，为多车通信提供有效手段。基于上述研究，项目组申请了国家发明专利4项，发表学术论文15篇，其中SCI检索论文10篇，EI检索论文5篇，获得省级科学奖励1项。在项目组的支持下，共有12名研究生参与研究，其中4名已顺利毕业。本项目的研究成果隶属于智能交通领域，涉及优化理论、区块链和人工智能技术，相关技术的研发将为智能交通和车联网技术的落地提供理论支撑。