Stochastic Data-Based Denial-of-Service Attack Strategy Design Against Remote State Estimation in Interval Type-2 T–S Fuzzy Systems

Abstract

This article focuses on designing a novel stochastic data-based denial-of-service (DoS) attack strategy that can intercept measurements and reduce the quality of remote state estimation in interval type-2 (IT2) T–S fuzzy systems. The new proposed DoS attack strategy has two main characteristics. First different from most of the existing DoS models, it is data-based, which mainly attacks the packets playing important roles in the system performance. Therefore, compared with the existing indiscriminate DoS attack approaches, the data-based DoS attack is more intelligent and can cause larger disruptions. Second, the proposed attack strategy is random, which has better concealment than the periodic or consecutive DoS attack models. Then, by using a stochastic method, the relationship among attack effect, attack probability, and attack parameter are analyzed and shown in Theorem <xref ref-type="theorem" rid="theorem9" xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">1</xref> . Furthermore, if the attacks are energy-constrained, the relation between the attack rate threshold and the upper bound of the attack parameter is expressed analytically, which is shown in Theorem <xref ref-type="theorem" rid="theorem12" xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</xref> . Simulation results demonstrate that, compared with some existing DoS models, the proposed stochastic data-based DoS attack strategy brings about greater destructiveness to the estimation quality of the systems, which demonstrates the effectiveness of the proposed attack strategy from the attacker’s perspective.