Malicious Anomaly Detection Approaches Robustness in Manufacturing ICSs

For the past decade, manufacturing Industrial Control Systems (ICSs) have suffered from targeted attacks against their physical system and control integrity, resulting in financial material losses. Among protective answers to this malicious threat, Anomaly Detection (ADS) based on behavioral models of ICS are highly regarded for ability detect zero-day attacks. However, design accurate non-obsolescent detection is not as an easy task a constantly changing environment. Thus, paper provides overview ADSs flaws issued unpredictable management its heterogeneous Behavioral will be introduced light four attributes: method, modeled behavior, lifecycle model nature. Then, each these attributes discussed regard robustness different environmental factors uncertainties they affected by.