SCVIC-TS-2022: Network intrusion data with original raw network packets

To apply TS-NFM, a network intrusion dataset must have original raw network packets rather than extracted features and complete labeling information; consequently, this dataset uses the CIC-IDS-2017 dataset (I. Sharafaldin, A. Habibi Lashkari, and A. A. Ghorbani, “Toward Generating a New Intrusion Detection Dataset and Intrusion Traffic Characterization:,” in Proc. of the 4th Intl. Conf. on Information Systems Security and Privacy, 2018). The raw network packets (PCAP format) are fed into the  Time Series Network Flow Meter (TS-NFM) proposed in our paper*  with a time window of two minutes (same to the extracted features from CIC-IDS-2017), resulting in the SCVIC-TS-2022 dataset. MTS's maximum length L is 511,681 due to the network's fast speed. The number of features/dimensions d is thirteen, which includes the direction of a packet, IAT, size in bytes, and ten TCP flags. *If you are using the SCVIC-TS-2022 dataset, plesae cite the following paper:J.Liu, M.Simsek, M. Nogueira, B. Kantarci, Multidomain transformer-based deep learning for early detection of network intrusion, IEEE Global Communications Conference (Globecom), Kuala Lumpur, Malaysia, pp. 1-6, Decembe 2023.