MINIDS: Multi-domain In-Vehicle Network Intrusion Detection Dataset for SCIMI

- This is the dataset of our paper "SCIMI: Semantic-Aware Collaborative Intrusion Detection for Multi-domain In-Vehicle Networks", which is accepted by 63rd ACM/IEEE Design Automation Conference (DAC 2026). - This repository is the public version of our dataset. - We conducted multi-stage attack experiments on a real autonomous transport vehicle and collected traces from the CAN, Ethernet, and host systems. - The label.txt file contains annotations of the attack periods, which are synchronized with the timestamps in all data. Each attack's can.csv file includes frame-level labels. Metrics files contain the host log of the vehicle. Output.pcap is the recorded Ethernet traffic. - You can find more details about this dataset in our paper. - Anyone who uses this dataset in his/her work should cite our paper. - Download: dac_pub_dataset.zip dac_pub_dataset.z01 dac_pub_dataset.z02 dac_pub_dataset.z03 - Run: unzip dac_pub_dataset.zip