CBTCset: A Reference Dataset for Detecting Misbehavior Attacks in CBTC Networks

Communication-Based Train Control (CBTC) systems are automatic train control systems widely used in railway networks. They are based on wireless bidirectional communications between trains and wayside units (WSUs). WSUs send operational commands to trains to adjust their mobility and ensure the safety and efficiency of operations. CBTC systems are safety critical with stringent reliability requirements. Enhancing their reliability involves securing communications against cyber attacks. This work presents CBTCset, a dataset to detect misbehavior attacks in CBTC networks. These attacks represent reliability threats since they can cause collisions and derailments. CBTCset addresses the gap represented by the absence of public reference datasets to detect attacks in CBTC networks. The dataset is publicly available for researchers in industry and academia on GitHub
¹
. We provide it as an open source to allow researchers to contribute new attacks and future extensions. This can facilitate comparable evaluation of detection techniques besides reproducibility and validation of results.