Search

The study focuses on the design of a diagnosis tool to detect a content modification attack of an uncertain network control system. In the literature, a content modification attack is designed to produce trajectory modification on tele-operated systems by modifying data transmitted through a network. Attack is difficult to detect because modification is transparent to classical detection tools. Moreover, uncertainties in this type of system complexify detection. So, an interval finite memory observer is used to estimate this type of system with uncertainties. Robust estimations provided by finite memory observers are combined with a specific strategy to detect modifications to measurements, further to an attack. Moreover, the corrupted data are corrected using an estimator to let the system continue working. A simulation of two haptic bilateral tele-operated systems validates the detection and data integrity correction strategies developed in this work. [ABSTRACT FROM AUTHOR]