Mean-square exponential stabilization of memristive neural networks: Dealing with replay attacks and communication interruptions

Shuai Xiao, Zhen Wang*, Xindong Si, Gang Liu

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

Abstract

This paper investigates the mean-square exponential stabilization (MSES) of memristive neural networks (MNNs) under replay attacks and communication interruptions. The research will revolve around the following two questions: Firstly, facing replay attacks and communication interruptions, how to design an appropriate controller? Secondly, how to ensure the MSES of MNNs under higher replay attack rate and communication interruption rate? To address these challenges, a novel intermittent sampled-data control scheme is established by considering the mathematical characteristics of replay attacks and communication interruptions. Furthermore, interval-dependent Lyapunov functions are constructed based on the Lyapunov stability theory and inequalities techniques, and two sufficient criteria for MSES of MNNs under the mentioned risks are derived. Thus, the control gain is obtained by solving a series of linear matrix inequalities. In addition, two algorithms are designed to investigate the maximum allowable replay attack rate and communication interruption rate for the system, respectively. Finally, two numerical examples are given to verify the effectiveness of the proposed scheme.

Original languageEnglish
Article number108188
JournalCommunications in Nonlinear Science and Numerical Simulation
Volume138
DOIs
Publication statusPublished - Nov 2024
Externally publishedYes

Keywords

  • Communication interruptions
  • Intermittent sampled-data control
  • Mean-square exponential stabilization
  • Memristive neural networks
  • Replay attacks

Fingerprint

Dive into the research topics of 'Mean-square exponential stabilization of memristive neural networks: Dealing with replay attacks and communication interruptions'. Together they form a unique fingerprint.

Cite this