The increasing use of the Internet in vehicles has made travel more convenient. However, there are several technological weaknesses that intelligent vehicle attackers can exploit, creating several security issues. These security concerns have drawn attention to the safety protection technologies of the in-vehicle system. We investigated the application of deep learning-based recurrent neural networks (RNNs) for enhanced auto encoder networks and intrusion detection in automobiles. To effectively comprehend the vehicle’s boundary behaviour and identify invasive behaviour, we merged two algorithms. The correctness and efficacy of the proposed model were verified through an analysis of actual vehicle data. The experiment’s results show that integrating the two technologies can successfully and dependably identify aberrant border behaviour. The model’s parameters are updated iteratively using the time-based backpropagation technique. We verified that the suggested model in the study may attain a detection accuracy of more than 96%.
An Intrusion System for Vehicular Ad Hoc Networks Using Deep Learning
Studies Comp.Intelligence
2025-06-09
29 pages
Article/Chapter (Book)
Electronic Resource
English