A Sensor Watermarking Design for Threat Discrimination

This paper proposes a sensor additive switching watermark methodology for detecting physical faults and replay attacks, then identifying the occurring threat type: either fault or attack. The includes generator in plant side remover control monitoring side. protocols of their common seeds are specifically designed, guaranteeing switch synchronization both nominal cases, allowing asynchronization attack case. latter is used discriminating two considered cases. detectability discrimination ability proposed approach, characterizing class attacks that can be detected discriminated, rigorously investigated. A simulation example presented to illustrate effectiveness approach.