Ultimate Boundedness Control for Networked Singularly Perturbed Systems With Deception Attacks: A Markovian Communication Protocol Approach

In this study, the ultimate boundedness controlfor a type of networked singularly perturbed systems (SPSs) with communication constraints and deception attacks is explored. To improve the observer performance, the measurement outputs are quantized with the aid of a logarithmic quantizer. Meanwhile, the Markovian communication protocol (MCP) is forwarded to schedule the transmission sequence of the quantized signals. Unlike the conventional MCP, a novel nonhomogeneous MCP is proposed to further alleviate the communication bandwidth usage, whose transition probabilities are time-varying. By virtue of the hidden Markov model, the mismatches between the transmission mode over nonhomogeneous MCP and its detected ones are revealed, and an asynchronous observer-based controller is formed. Then, by establishing a novel singular-perturbation-parameter-based polytope Lyapunov-Krasovskii functional, the ultimate boundedness of the augmented networked SPS with randomly occurring deception attacks is elicited, and the local minimization of the controlled output is guaranteed. Finally, the effectiveness and applicability of the propounded control strategy are validated by two simulation examples.

Automated summary

This study explores ultimate boundedness control for a type of networked singularly perturbed systems with communication constraints and deception attacks. A logarithmic quantizer is used to improve observer performance, and a novel nonhomogeneous MCP is proposed to reduce communication bandwidth usage. By establishing a polytope Lyapunov-Krasovskii functional based on singular perturbation parameters, the ultimate boundedness of the networked SPS with deception attacks is ensured, and controlled output minimization is guaranteed. The effectiveness and applicability of the control strategy are validated through simulation examples.