<p>With the rapid development of intelligent connected vehicles and urban smart facilities, the Internet of Vehicles (IoV) has been applied more and more widely. However, the security of the IoV also faces huge challenges. In recent years, Digital Twin (DT) technology has been applied to the deployment and design of IoV. However, the use of DT technology to design intrusion detection systems (IDS) has not been widely explored. In the paper, we discuss the possibility of using DT for intrusion detection in the IoV and construct the DT of IoV called IoVDTNet. We simulate four intrusion attacks on IoVDTNet: command injection attack, Denial-of-Service attack, Scale tampering attack, and Random tampering attack, and develop an IDS based on optimal ensemble learning to protect the information security of IoVDTNet. Experimental results show that the IDS has an accuracy rate of 99.99<InlineEquation ID="IEq1"> <InlineMediaObject> <ImageObject Color="BlackWhite" FileRef="11227_2025_8044_Article_IEq1.gif" Format="GIF" Height="16" Rendition="HTML" Resolution="72" Type="Linedraw" Width="15" /> </InlineMediaObject> <EquationSource Format="TEX">$\%$</EquationSource> <EquationSource Format="MATHML"><math> <mo>%</mo> </math></EquationSource> </InlineEquation> in the CAN-Hacking dataset representing the intra-vehicle networks and 99.91<InlineEquation ID="IEq2"> <InlineMediaObject> <ImageObject Color="BlackWhite" FileRef="11227_2025_8044_Article_IEq1.gif" Format="GIF" Height="16" Rendition="HTML" Resolution="72" Type="Linedraw" Width="15" /> </InlineMediaObject> <EquationSource Format="TEX">$\%$</EquationSource> <EquationSource Format="MATHML"><math> <mo>%</mo> </math></EquationSource> </InlineEquation> in the CICIDS2017 dataset representing the external vehicle networks. In addition, the accuracy rate of the real-time intrusion dataset generated by IoVDTNet is 99.80<InlineEquation ID="IEq3"> <InlineMediaObject> <ImageObject Color="BlackWhite" FileRef="11227_2025_8044_Article_IEq1.gif" Format="GIF" Height="16" Rendition="HTML" Resolution="72" Type="Linedraw" Width="15" /> </InlineMediaObject> <EquationSource Format="TEX">$\%$</EquationSource> <EquationSource Format="MATHML"><math> <mo>%</mo> </math></EquationSource> </InlineEquation>, and the average detection time of each data packet is less than 1 ms. This study demonstrates the feasibility and practicality of using DT technology to protect IoV.</p>

An intrusion detection system for Internet of Vehicles based on digital twin

期刊论文
发表日期：

2025年11月14日

查看全文

Yingqing Wang,
Guihe Qin,
Minghui Sun,
Yanhua Liang,
Xuezhu Yang,
Muxi Li,
Chuang Hu

摘要

With the rapid development of intelligent connected vehicles and urban smart facilities, the Internet of Vehicles (IoV) has been applied more and more widely. However, the security of the IoV also faces huge challenges. In recent years, Digital Twin (DT) technology has been applied to the deployment and design of IoV. However, the use of DT technology to design intrusion detection systems (IDS) has not been widely explored. In the paper, we discuss the possibility of using DT for intrusion detection in the IoV and construct the DT of IoV called IoVDTNet. We simulate four intrusion attacks on IoVDTNet: command injection attack, Denial-of-Service attack, Scale tampering attack, and Random tampering attack, and develop an IDS based on optimal ensemble learning to protect the information security of IoVDTNet. Experimental results show that the IDS has an accuracy rate of 99.99 $\%$ $%$ in the CAN-Hacking dataset representing the intra-vehicle networks and 99.91 $\%$ $%$ in the CICIDS2017 dataset representing the external vehicle networks. In addition, the accuracy rate of the real-time intrusion dataset generated by IoVDTNet is 99.80 $\%$ $%$ , and the average detection time of each data packet is less than 1 ms. This study demonstrates the feasibility and practicality of using DT technology to protect IoV.