【作者】 刘小波；

【导师】 刘虹；

【作者基本信息】 华东师范大学 ， 工程硕士（专业学位）， 2022， 硕士

【摘要】 截至2021年6月,全国机动车保有量达3.84亿辆,其中汽车2.92亿辆,与此同时由机动车造成的交通事故和环境污染引起社会广泛关注。根据国家统计局发布的数据显示,交通事故造成的人员死亡已经成为我国非正常死亡的主要因素之一。车联网(Internet of Vehicles,IOV)作为智能交通系统(Intelligent Transportation System,ITS)的重要组成部分,能够有效的降低交通事故率、缓解城市拥堵。但是由于车联网具有开放性、移动性、实时性和拓扑结构动态变化等特点,使其容易受到各种各样的网络攻击。传统的加密、签名手段可以缓解车联网在认证、接入、数据传输等方面的安全问题,并不能解决车联网内部节点恶意攻击其他节点的问题。特别是在自动驾驶技术迅速发展的今天,一个内部节点向车联网中的其他节点发送虚假消息或者进行恶意攻击,可能导致其作出错误的决策,从而引发严重的交通事故,对司乘人员的生命财产安全构成威胁。因此需要在车联网中引入“信任管理”机制,对内部节点进行信任管理,动态识别出恶意节点,筛选出可信的节点进行协作。针对车联网中的“信任管理”机制,本文主要做了以下三个方面的工作:·设计一个分层信任管理框架根据车联网的特点,将数字孪生与边缘计算引入其中,然后设计一个分层信任管理框架,并详细定义了框架中的各个子系统和工作流程。·实现抗攻击信任评估算法实现一个具有抗攻击能力的信任评估算法,该算法部署到分层信任管理框架的各个子系统中,用于局部信任值和全局信任值的计算。·Veins平台进行仿真实验使用Veins软件对车联网通信进行仿真,按照分层信任管理框架定义的模块和流程设定算法验证程序,将仿真得到的数据用于算法验证,最后分析抗攻击信任评估算法在各种攻击场景下的表现。本文根据车联网的特点,构建一个全新的车载数字孪生边缘网络,然后基于该网络的结构设计了一个分层信任管理框架,最后实现了该框架对应的抗攻击信任评估算法。实验仿真结果表明,本文提出的抗攻击信任评估算法不仅能够有效的识别恶意节点,还能抵抗专门针对信任管理机制的攻击。更多还原

【Abstract】 As of June 2021,the number of motor vehicles in the country reached 384 million,including 292 million cars,while traffic accidents and environmental pollution caused by motor vehicles have caused widespread social concern.According to the data released by the National Bureau of Statistics,deaths caused by traffic accidents have become one of the main factors of unnatural deaths in China.On the one hand,Internet of Vehicles(IOV),as an important part of Intelligent Transportation System(ITS),can effectively reduce traffic accident rate and ease urban congestion.On the other hand,due to the characteristics of openness,mobility,real-time and dynamic change of topology,IOV are vulnerable to various network attacks.Although the traditional encryption and signature means can alleviate the security problems of IOV in authentication,access and data transmission,they do not solve the problem of malicious attacks on other nodes by internal nodes of IOV.Especially with the rapid development of autonomous driving technology today,an internal node sending false messages or carrying out network attacks to other nodes in the IOV,leading to wrong decisions,may cause serious traffic accidents and pose a threat to the lives and properties of drivers and passengers.Therefore,it is necessary to introduce ”trust management” mechanism in IOV to manage the trust of internal nodes,to dynamically identify malicious nodes and filter out trusted nodes for collaboration.For the ”trust management” mechanism in IOV,this paper focuses on the following aspects.· Design a hierarchical trust management framework According to the characteristics of the IOV,digital twins and edge computing are introduced into the IOV,and then a hierarchical trust management framework is designed,and each subsystem and workflow in the framework is defined in detail.· Implementation of an anti-attack trust evaluation algorithm Implement an anti-attack trust evaluation algorithm that is deployed to each subsystem of the hierarchical trust management framework for the calculation of local trust values and global trust values.· The Veins platform was used for simulation experiments Simulation of the IOV communication using Veins software,verification of the algorithm using the data obtained from the simulation,and finally analysis of the performance of the anti-attack trust management algorithm in various attack scenarios.In this paper,we construct a new vehicle-mounted Digital Twin edge Network based on the characteristics of IOV,then design a hierarchical trust management framework based on this network,and finally implement the trust evaluation algorithm corresponding to this framework.Experimental simulation results show that the anti-attack trust evaluation algorithm proposed in this paper is not only effective in identifying malicious nodes,but also resistant to attacks specifically targeting the trust management system.更多还原