A Fully Labeled Multi Factor Dataset for Anomalous State Detection in a Radio Access Network

In 2023 and 2024, two experiments were conducted to investigate anomalous state detection
in a commercial off-the-shelf radio access network (RAN) within a controlled laboratory
environment. The RAN configurations were systematically varied by adjusting modulation
schemes, payload types, and traffic volumes, while capturing responses from multiple
layers of the RAN architecture. For each configuration, responses were collected in two
distinct encryption states: a baseline state with Advanced Encryption Standard (AES) encrypted
data and an anomalous state with encryption disabled by turning off the user equipment
ciphering capability at the base station. The first experiment employed a conducted
propagation channel, while the second utilized a radiated propagation channel in an anechoic
chamber. The labeled dataset generated from both experiments is publicly available
and can be used for further research in network anomaly detection, performance monitoring,
and security. Detailed documentation on the experiment design, data formatting, and
methodology is provided.