A Masking Protocol for Nonlinear and Incrementally Passive Average Consensus Algorithms

Abstract

This letter introduces a masking protocol to enhance the security of a consensus protocol for nonlinear multi-agent systems. The proposed approach involves adding a masking signal to each agent’s output and applying a de-masking filter at the receiving agent. We establish sufficient conditions to ensure that the proposed security protocol preserves output consensus. Furthermore, numerical simulations demonstrate its effectiveness in preventing eavesdropping and false data injection attacks.