Preprocessed VeReMi Dataset for IoV Intrusion Detection

This dataset contains a preprocessed and balanced version of the VeReMi Dataset for Misbehavior Detection in Vehicular Ad-Hoc Networks. The original dataset suffers from severe class imbalance and redundant features, making it challenging for machine learning applications in Intrusion Detection Systems (IDS) and Internet of Vehicles (IoV) research. To address these issues, the dataset has been: Downsampled (10%) for computational efficiency (~724MB). Balanced using synthetic benign data generation. Feature-selected to retain critical attributes like rcvTime, pos_0, pos_1, AttackerType, etc. Filtered using rule-based classification for benign sample identification. This dataset is suitable for anomaly detection, intrusion detection, and vehicular security research. Original Dataset Source: VeReMi Dataset on KagglePreprocessed Data DOI/Link: 10.5281/zenodo.14903687