5G-JA-UAD 5G Jamming Attack – Under Attack Dataset

The <i>5G_SA_Jamming</i> dataset is designed to support the development of security mechanisms in 5G standalone (SA) networks. It includes a total of 46,816 labeled frames captured in a controlled private 5G SA environment, of which 3,464 correspond to traffic affected by jamming attacks and the rest to normal or legitimate disconnection events. The dataset was generated using real 5G equipment and includes multiple attack sessions carried out with a programmable jammer operating on the 3319.68 MHz band.Traffic traces focus on core signaling and user plane protocols such as NGAP, SCTP, GTP, and TCP, capturing anomalies induced by interference. The data is intended for the training and evaluation of machine learning models, particularly for intrusion detection systems (IDS) in modern 5G deployments, both in research and industrial settings.