Distributed filtering for nonlinear systems under Dempster–Shafer theory subject to malicious cyber attacks

Abstract

This paper addresses the problem of network security for distributed filtering when subject to malicious cyber attacks. First, based on the Dempster-Shafer (D-S) theory a protector for each sensor is proposed, that can decide whether to use data sent by neighboring sensors. Furthermore, a distributed estimator is proposed for the nonlinear system with the designed detector. Besides, a sufficient condition is presented for the stability of the estimator equipped with the designed protector when under cyber attacks. Finally, several numerical examples are presented to verify the proposed protection strategy's efficacy.

LPWAN Cyber Security Risk Analysis: Building a Secure IQRF Solution

“…Although there are several steps forming the fundamental process for most of cybersecurity risk assessment approaches, there is still a degree of flexibility in how these steps are implemented. Thus, this has resulted in the emergence of various traditional approaches for this purpose such as fault tree analysis [ 18 ], event tree analysis [ 19 ], decision matrix risk assessment [ 20 ], weighted risk analysis [ 21 ], fuzzy set theory [ 22 ], Dempster–Shafer theory [ 23 ], game theoretic computing [ 24 ], cyber security game [ 25 ], operationally critical threat asset and vulnerability evaluation (OCTAVE) [ 26 ], and FMEA [ 27 ]. Before the development of the IoT and WSNs, the use of these approaches was mostly preferred to manage security risks.…”