Simultaneous Detection and Characterization of Time Delay Attack

This technical report discusses the simultaneous detection and characterization of time delay attacks (TDA) in cyber-physical systems (CPS). It begins by outlining the background of CPS, including the roles of actuators and sensors, and the vulnerabilities associated with communication channels. The report details the nature of TDA, explaining how malicious delays can be introduced without tampering with the signals. It presents a system model for assessing the impact of TDA, including the challenges faced in real-time versus post-mortem analysis. The authors propose a solution utilizing a hierarchical long short-term memory (HLSTM) architecture to enhance detection and characterization capabilities. Evaluation metrics such as mean absolute error (MAE) and accuracy are used to compare the proposed model against baseline models, demonstrating its effectiveness in reducing reaction latency and improving detection accuracy. The report concludes with insights on the implications of TDA and the necessity for robust detection mechanisms in CPS.