A Separation Signal for Heterogeneous Networks

Abstract

Organizing a large-scale, heterogeneous network of agents into clusters based on the agents’ class is a useful preprocessing step for cooperative tasks, where agents with the same capabilities need to be in the same location to cooperatively solve a task. In this paper, we investigate whether it is possible to apply an exogenous control signal to a heterogeneous network of agents, such that these agents form clusters of the same class of agents. We demonstrate that if each agent belongs to one of M “weight” classes and executes a weighted, forced agreement protocol, then it is possible to apply an external input signal that separates the network into M clusters corresponding to the M classes of agents.