Study of Adaptive Spoofing Attack Detection in Connected Vehicles with Dilated and Attention-Driven Neural Networks

• Version
• Download 43
• File Size 694.91 KB
• File Count 1
• Create Date 18 June, 2025
• Last Updated 12 July, 2025

Authors : Vishal R. Deshmukh, Prof. Dr. Indrabhan S. Borse   , Dr. Balveer Singh

DOI : 10.46335/IJIES.2025.10.8.11

Abstract – The rapid evolution of Intelligent Transportation Systems (ITS) has significantly enhanced vehicular communication, fostering safer and more efficient road networks. However, the growing interconnectivity exposes vehicular networks to a variety of cyberattacks, with spoofing attacks emerging as a critical threat. These attacks allow malicious actors to impersonate legitimate vehicles or infrastructure nodes, jeopardizing network security and road safety. This study proposes an adaptive spoofing attack detection framework utilizing a hybridized neural network model that combines Dilated Convolutional Neural Networks (DCNNs) and Attention Mechanisms. The DCNN component captures multi-scale spatial and temporal dependencies in vehicular data, ensuring broad context awareness, while the attention mechanism dynamically prioritizes crucial features, enhancing detection accuracy. In real-time, leveraging both vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) communications to analyze incoming signals and detect anomalies. The adaptive nature of the model allows it to adjust to varying traffic patterns and evolving attack strategies. Extensive experiments using simulated vehicular network datasets demonstrate the model's effectiveness, achieving high detection rates and reduced false positives compared to traditional methods. The study also explores the computational efficiency of the framework, ensuring its feasibility for deployment in resource-constrained vehicular environments. Ultimately, this research contributes to strengthening cybersecurity in connected vehicle ecosystems by presenting a robust, intelligent, and adaptable solution for spoofing attack detection. The findings underscore the potential for integrating advanced AI techniques into ITS, paving the way for more secure and resilient vehicular networks.