Privacy-Preserving Leader-Following Consensus via Finite Time-Varying Transformation
Yaqi Wang, Wei Wang · 2024
This paper explores the privacy protection of leader-following consensus in multi-agent systems (MASs). It is widely recognized that the conventional protocol for leader-following consensus may result in the exposure of sensitive information when agents exchange initial states with their neighbors. To conquer this problem, we introduce a time-varying transformation function to ensure privacy protection for the following agents in the system. The purpose of adopting the transformation function is is to make the initial states of the followers indistinguishable, effectively masking their true states in asymptotic time. The paper provides a proof that both leader-following consensus and privacy protection for the following agents can be achieved using this approach. Additionally, numerical simulation proves that the leader-following consensus of multi-agent systems can be achieved under this method.