AFTN Message Anomaly Detection Sample Dataset

Constructed based on the civil aviation message switching system, it comprises standard message samples and perturbed message samples, organized by positive and negative categories to support studies in classification, anomaly detection, and robustness evaluation. The data covers multiple operational scenarios, including departure, landing, flight plan changes, and system status, reflecting structural and temporal characteristics under real-world operating conditions.

To ensure data security, sensitive fields such as AFTN addresses, airport codes, and flight numbers have been consistently pseudonymized, preserving structural relationships and statistical characteristics while mitigating the risk of information disclosure. The dataset is provided in CSV format and is suitable for research in aviation message governance and domain-specific security.