Research Article
Deep Learning-Based Traffic Information Prediction Methods in the Internet of Vehicles
@INPROCEEDINGS{10.1007/978-3-031-86203-8_5, author={Chenguang He and Bohan Zhang and Liang Ye and Hua Tan}, title={Deep Learning-Based Traffic Information Prediction Methods in the Internet of Vehicles}, proceedings={Wireless and Satellite Systems. 14th EAI International Conference, WiSATS 2024, Harbin, China, August 23--25, 2024, Proceedings, Part II}, proceedings_a={WISATS PART 2}, year={2025}, month={3}, keywords={Internet of Vehicles Deep Learning Neural Network Traffic Information Prediction Spatial and Temporal Dependence}, doi={10.1007/978-3-031-86203-8_5} }- Chenguang He
Bohan Zhang
Liang Ye
Hua Tan
Year: 2025
Deep Learning-Based Traffic Information Prediction Methods in the Internet of Vehicles
WISATS PART 2
Springer
DOI: 10.1007/978-3-031-86203-8_5
Abstract
In recent years, the rapid development of deep learning technology has provided many methods to predict traffic information prediction. In the Internet of Vehicles (IoV), accurate and real-time traffic information prediction plays an important role in improving system performance and user experience. How to effectively capture the temporal and spatial dependencies of traffic information is a major challenge in this field. In this paper, we focus on three neural network models (GRU, TGCN and TGCN-att) for traffic information prediction and train these three models using real datasets. We analyzed the outputs of each model separately, compared the performance metrics such as mean square error (RMSE) and mean absolute error (MAE) between the predicted and real values, and calculated the accuracy of the predictions of each model. The simulation results show that since road networks generally have a complex topology, correctly capturing the spatial dependence between data is very important for improving the prediction accuracy of the models when performing traffic information prediction.
