AegisSat: Telemetry dataset of a physical CubeSat environment with labeled attacks

We used the testbed to create a comprehensive dataset, the main purpose of which is to support further research on satellites in the cybersecurity domain. The dataset collected for the evaluation presented in this paper comprises data from hundreds of simulations designed to emulate satellite operations in space. The simulations range from one lap, which is 90 minutes long, to a few hours, and approximately 5,000 records are generated per lap. The dataset consists of Satellite telemetry data from onboard sensors, recorded at a frequency of one reading per second Details on commands issued from the ground station; and Records corresponding to attack scenarios, which are explicitly labeled with the attack type and associated parameters. Each sample contains general information, including the simulation ID, timestamp, satellite name, and telemetry data, with other fields available to store relevant details when operational commands are issued and attacks are simulated.The telemetry data includes readings from sensors monitoring various components of the satellite. The components we collect data on are: TCS: provides information on the TCS, including its operational mode and temperature Payload: captures data on the camera status, CPU usage, and RAM usage percentage; (c) \textbf{EPS:} captures readings from sensors in the satellite's EPS OBC: contains data from the OBC, such as processor temperature and RAM utilization data. When attacks are simulated in a given sample, a field describing the names of the attacks and relevant parameters is populated.