Securing intelligent transportation systems in the internet of vehicles through an AI-driven intrusion detection framework
Abstract
The integration of the Internet of Things (IoT) with modern vehicles has transformed transportation systems into intelligent, interconnected Internet of Vehicles (IoV) networks. However, this integration increases the attack surface, exposing IoV networks to threats that compromise operational reliability and passenger safety. This work introduces a configurable simulation platform for designing and evaluating IoV networks across diverse topologies and network parameters under normal and adversarial conditions. The platform supports multiple Distributed Denial-of-Service (DDoS) attacks across the communication stack, including application-layer Slowloris and HTTP, transport-layer SYN and UDP, and reflection-based DNS and NTP amplification. It incorporates a dataset-generation module that captures IoV behavior under benign and adversarial conditions to support training and evaluation of security models. Using these datasets, an AI-based detection framework is developed using machine learning, deep learning, and transformer models. A multi-layer evaluation combining communication and AI classification metrics demonstrates the platform’s capability to model complex IoV behavior and support security methods. It further measures deployment trade-offs involving inference latency and communication-energy cost, which reflect the practical constraints of resource‑limited devices. Experimental results show that amplification attacks are the most destructive, while cellular topology demonstrates the highest resilience, with 56.5% loss, whereas VANET and Manhattan show 87.5% loss. Under edge constraints, transformer models achieve 92–93% recovery but incur 136–181 ms inference latency, deep learning achieves 89–91% recovery with 17–45 ms latency, and machine learning provides 72–87% recovery with the lowest latency of 12–27 ms, highlighting the trade-off between detection capability and computational overhead.
Related
- Generative Self-Attention for Adaptive Real-Time Traffic Event Detection in Next-Generation Intelligent Transportation Systems
- Intelligent defense at the edge: a comprehensive survey of federated learning, TinyML and explainable AI for intrusion detection in IoT and IIoT ecosystems
- TRIDENT-Risk: An AI-driven temporal ensemble framework for proactive high-severity accident risk forecasting in Intelligent Transportation Systems
- A resource-aware framework for energy and bandwidth efficient IoT intrusion detection using lightweight federated learning over SDN
- A generative adversarial network-augmented dual-stream Mamba-transformer framework for advanced intrusion detection and cyber threat mitigation in smart healthcare networks
BibTeX
@article{alshammari2026securing,
title = {{Securing intelligent transportation systems in the internet of vehicles through an AI-driven intrusion detection framework}},
author = {Sajjad Salim Obaid Alshammari and Mina Malekzadeh},
journal = {Scientific Reports},
year = {2026},
month = oct,
doi = {10.1038/s41598-026-77083-7},
url = {https://www.nature.com/articles/s41598-026-77083-7},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Securing intelligent transportation systems in the internet of vehicles through an AI-driven intrusion detection framework
AU - Sajjad Salim Obaid Alshammari
AU - Mina Malekzadeh
PY - 2026
DA - 2026/10/24
JO - Scientific Reports
DO - 10.1038/s41598-026-77083-7
UR - https://www.nature.com/articles/s41598-026-77083-7
AB - The integration of the Internet of Things (IoT) with modern vehicles has transformed transportation systems into intelligent, interconnected Internet of Vehicles (IoV) networks. However, this integration increases the attack surface, exposing IoV networks to threats that compromise operational reliability and passenger safety. This work introduces a configurable simulation platform for designing and evaluating IoV networks across diverse topologies and network parameters under normal and adversarial conditions. The platform supports multiple Distributed Denial-of-Service (DDoS) attacks across the communication stack, including application-layer Slowloris and HTTP, transport-layer SYN and UDP, and reflection-based DNS and NTP amplification. It incorporates a dataset-generation module that captures IoV behavior under benign and adversarial conditions to support training and evaluation of security models. Using these datasets, an AI-based detection framework is developed using machine learning, deep learning, and transformer models. A multi-layer evaluation combining communication and AI classification metrics demonstrates the platform’s capability to model complex IoV behavior and support security methods. It further measures deployment trade-offs involving inference latency and communication-energy cost, which reflect the practical constraints of resource‑limited devices. Experimental results show that amplification attacks are the most destructive, while cellular topology demonstrates the highest resilience, with 56.5% loss, whereas VANET and Manhattan show 87.5% loss. Under edge constraints, transformer models achieve 92–93% recovery but incur 136–181 ms inference latency, deep learning achieves 89–91% recovery with 17–45 ms latency, and machine learning provides 72–87% recovery with the lowest latency of 12–27 ms, highlighting the trade-off between detection capability and computational overhead.
SN - 2045-2322
ER -