Privacy-Preserving Consensus of Double-Integrator Multi-Agent Systems With Input Constraints

Abstract

Consensus is one of the most important topics in distributed multi-agent systems (MAS). In general, existing consensus approaches aim at driving agents to reach an agreement via negotiating with their local neighbors, which means that explicit state information is exchanged among agents. This leads to privacy breach. Thus, if agents' state information is important and sensitive, privacy preservation should be taken into account. In this paper, we propose a near-optimal consensus algorithm for double-integrator MAS under an undirected connected topology, which guarantees convergence, compliance with input constraints and privacy preservation of the agents in a distributed manner. By combining partial homomorphic cryptography with interaction dynamics, state information of agents can be well protected from honest-but-curious adversaries and external eavesdroppers. The privacy preserving property is proved via theoretical analysis, and the effectiveness of the algorithm is verified via computer simulations.