The content discusses achieving consensus in multi-agent systems under intermittent malicious attacks or failures. It introduces a detection algorithm to mitigate attacks, ensuring legitimate agents reach an agreement. The analysis shows that misclassification probabilities decrease over time, leading to almost sure convergence among legitimate agents. The deviation from the nominal consensus value is bounded, with a detailed explanation of the consensus dynamics and the impact of malicious agents.
Til et andet sprog
fra kildeindhold
arxiv.org
Vigtigste indsigter udtrukket fra
by Sarp... kl. arxiv.org 03-27-2024
https://arxiv.org/pdf/2403.17907.pdfDybere Forespørgsler