Dynamic resilient graph games for state-dependent jamming attacks analysis on multi-agent systems

Yurid Nugraha, Ahmet Cetinkaya, Tomohisa Hayakawa, Hideaki Ishii, Quanyan Zhu

Research output: Contribution to journalConference articlepeer-review

Abstract

A cybersecurity problem for a multi-agent consensus problem is investigated through a dynamic game formulation. Specifically, we consider a game repeatedly played between a jamming attacker and a defender. The attacker attempts to jam the links between a number of agents to delay their consensus. On the other hand, the defender tries to maintain the connection between agents by attempting to recover some of the jammed links with the goal of achieving faster consensus. In each game, the players decide which links to attack/recover and for how long to continue doing so based on a Lyapunov-like function representing the largest difference between the states of the agents. We analyze the subgame perfect equilibrium of the game and obtain an upper bound of the consensus time that is influenced by the strategies of the players. The results are illustrated with a numerical example.

Original languageEnglish (US)
Pages (from-to)3421-3426
Number of pages6
JournalIFAC-PapersOnLine
Volume53
Issue number2
DOIs
StatePublished - 2020
Event21st IFAC World Congress 2020 - Berlin, Germany
Duration: Jul 12 2020Jul 17 2020

Keywords

  • Consensus problem
  • Cybersecurity
  • Game theory
  • Multi-agent systems

ASJC Scopus subject areas

  • Control and Systems Engineering

Fingerprint Dive into the research topics of 'Dynamic resilient graph games for state-dependent jamming attacks analysis on multi-agent systems'. Together they form a unique fingerprint.

Cite this