Publicação

Reputation-based resilient consensus with privacy guarantees

Ver documento

Detalhes bibliográficos
Resumo:Reputation-based consensus methods have gained significant attention in distributed systems to mitigate the impact of malicious agents and ensure reliable decision-making. However, privacy concerns arise when sensitive information is shared among participants. In this paper, we propose a reputation-based consensus method that incorporates resilience and privacy guarantees, providing a balance between preserving participants' privacy and maintaining the accuracy of the resilient consensus process. Our approach leverages the notion of privacy to add noise to the agents' states, ensuring that individual contributions are masked while still enabling the detection of malicious behavior. We present a detailed analysis of the privacy-accuracy trade-off and demonstrate the effectiveness of our method through simulations. The results show that our reputation-based consensus method with privacy guarantees offers robustness against attacks while preserving the privacy of participants, making it a promising solution for privacy-conscious resilient distributed systems.
Autores principais:Ramos, Guilherme
Outros Autores:Pequito, Sergio; Silvestre, Daniel
Assunto:Average consensus Consensus Privacy Reputation systems Resilient consensus Control and Systems Engineering Signal Processing Computer Networks and Communications Control and Optimization
Ano:2025
País:Portugal
Tipo de documento:artigo
Tipo de acesso:acesso embargado
Instituição associada:Universidade Nova de Lisboa
Idioma:inglês
Origem:Repositório Institucional da UNL
Descrição
Resumo:Reputation-based consensus methods have gained significant attention in distributed systems to mitigate the impact of malicious agents and ensure reliable decision-making. However, privacy concerns arise when sensitive information is shared among participants. In this paper, we propose a reputation-based consensus method that incorporates resilience and privacy guarantees, providing a balance between preserving participants' privacy and maintaining the accuracy of the resilient consensus process. Our approach leverages the notion of privacy to add noise to the agents' states, ensuring that individual contributions are masked while still enabling the detection of malicious behavior. We present a detailed analysis of the privacy-accuracy trade-off and demonstrate the effectiveness of our method through simulations. The results show that our reputation-based consensus method with privacy guarantees offers robustness against attacks while preserving the privacy of participants, making it a promising solution for privacy-conscious resilient distributed systems.