Intermittent control of delayed positive T-S fuzzy systems under dual-channel hidden DoS attacks

Abstract Modern networked control systems typically rely on multi-channel transmissions, including sensor-to-controller and controller-to-actuator channels. Attackers can achieve more stealthy and efficient disruption by simultaneously compromising dual channels, potentially rendering traditional single-channel defense strategies ineffective. Hidden denial-of-service attacks, a class of switched DoS attacks, involve stochastic partial blockages and time-varying attack intensities, thereby posing a more realistic threat scenario. To address this issue, this paper investigates the control problem of time-delayed positive Takagi-Sugeno fuzzy systems subject to dual-channel hidden denial-of-service attacks. Two mutually independent Markov processes are employed to characterize the composite stochastic incompleteness of information induced by the dual-channel attacks. A control framework based on a fuzzy observer is designed to mitigate the impact of state information not being safely utilized for feedback control. The proposed control strategy employs a state-dependent aperiodic intermittent operation, dynamically selecting between Active mode, Hold mode, and Dormant mode to minimize unnecessary control actuation. Sufficient conditions for mean-square exponential stability under the proposed controller are derived. Numerical simulations and the Lotka-Volterra population model demonstrate the effectiveness of the method in mitigating the effects of bounded dual-channel attacks on positive Takagi-Sugeno fuzzy systems.