Average Consensus of Whole-Process Privacy Preservation

Dear Editor, This letter introduces a novel algorithm for privacy preservation designed to safeguard both the initial and real-time states of agents under complete distributed average consensus. It addresses a gap in existing privacy preservation approaches that predominantly focus on protecting the initial state, with limited consideration for privacy implications throughout the entire process. The algorithm ensures the privacy of both the initial and real-time states by introducing perturbations to the consensus process, allowing agents to freely define these perturbations themselves. Additionally, the perturbations defined by agents arbitrarily do not compromise the accuracy of the consensus result. One of the main results derived is that no agent has access to the real-time state of another agent.