Optimal innovation-based deception attack on remote state estimationShow others and affiliations
2017 (English)In: Proceedings of the American Control Conference, Institute of Electrical and Electronics Engineers Inc. , 2017, p. 3017-3022Conference paper, Published paper (Refereed)
Abstract [en]
The security issue in cyber-physical systems has attracted growing interests in the last decades. This paper considers how false data injection attack can degrade the estimation quality of a remote state estimation system. In this system, smart sensors measure a dynamic process and send preprocessed data through a communication network to a remote estimator to estimate the process. It is assumed that there are malicious attackers in the communication network, who are able to obtain and falsify all the data sent by the sensors. It is common that the remote estimator is equipped with a residue-based detector to detect potential attacks. We propose a class of deception attack and analyze its feasibility. We show that the proposed attack enables the attacker to inject false data into the remote estimator without being detected. We derive a criterion to judge the optimality of performance of this type of attack in the sense of maximizing the estimation error covariance. Furthermore, we find that a simple linear attack strategy, which flips the sign of intercepted signal, satisfies the optimality criterion. We present numerical examples to illustrate our theoretical results.
Place, publisher, year, edition, pages
Institute of Electrical and Electronics Engineers Inc. , 2017. p. 3017-3022
Keywords [en]
Cyber-physical systems (CPS), deception attack, remote state estimation, Cyber Physical System, Estimation, State estimation, Estimation errors, Estimation quality, False data injection attacks, Optimality criteria, Pre-processed data, Remote state estimations, Embedded systems
National Category
Control Engineering
Identifiers
URN: urn:nbn:se:kth:diva-216437DOI: 10.23919/ACC.2017.7963410ISI: 000427033303013Scopus ID: 2-s2.0-85027007249ISBN: 9781509059928 (print)OAI: oai:DiVA.org:kth-216437DiVA, id: diva2:1164036
Conference
2017 American Control Conference, ACC 2017, 24 May 2017 through 26 May 2017
Note
QC 20171208
2017-12-082017-12-082022-06-26Bibliographically approved