ABSTRACT The leaders play an instrumental role in the leader‐follower formation control networks, guiding the followers to a prescribed formation pattern through their specialized capabilities or access to task‐specific information. However, the unique attributes of the leaders also render them more susceptible to being identified by adversaries. In this paper, we propose a novel leader‐follower formation control algorithm to conceal the leaders' identities in the networks, complicating the task for the adversaries to distinguish the leaders from other agents. Specifically, the proposed algorithm consists of two primary design elements. First, additional noises are directly injected into the leaders, simulating the presence of agents influencing them. Second, virtual leaders are designed to guide the 1‐hop followers, further obfuscating the direct interaction between the leaders and the 1‐hop followers. Furthermore, this paper provides a comprehensive convergence analysis of the proposed algorithm to ensure its efficacy in achieving the formation task. Extensive simulations are conducted to validate the effectiveness of the proposed algorithm and demonstrate its performance in leadership protection against various leadership inference methods.
Wang et al. (Wed,) studied this question.