Robust decentralized diagnosability of networked discrete event systems against DoS and deception attacks. (May 2022)

• Record Type: Journal Article
• Authors: Alves, Marcos V.S.; Barcelos, Raphael J.; Carvalho, Lilian K.; Basilio, João C.
• Is Part Of: Nonlinear analysis. Volume 44(2022)
• Publication Date: 2022-05
• Subjects: Discrete event systems -- Automaton -- Fault diagnosis -- Robustness -- Denial-of-service and deception attacks; Nonlinear functional analysis -- Periodicals; Analyse fonctionnelle non linéaire -- Périodiques; Nonlinear functional analysis; Periodicals; 515.7248
• DOI: 10.1016/j.nahs.2022.101162
• Languages: English
• ISSNs: 1751-570X

Abstract

Abstract: Denial-of-Service (DoS) are attacks conducted by malicious agents that consists in disrupting, temporally or indefinitely, the services provided by a communication network. When a malicious agent gets access to some network node, it may also perform deception attacks by inserting valid packets with fake information into vulnerable channels. We address, in this paper, DoS and deception attacks (DoS-D attack) that flood some communication channels with fake packets causing delays, loss of observations and insertion of fake observations, and their implications in decentralized fault diagnosability of networked discrete event systems (NDES). To this end, we propose an automaton model for NDES subject to DoS-D attacks that represents the adverse effects of DoS-D attacks on the observations of local diagnosers. We introduce a new codiagnosability definition called DoS-D-robust codiagnosability, and present a necessary and sufficient condition for a language to be DoS-D-robustly codiagnosable. We also propose a verification algorithm for regular languages to check DoS-D-robust codiagnosability.