On Heterogeneity in Foraging by Ant-Like Colony: How Local Affects Global and Vice Versa

摘要

In this paper, we discuss influence of heterogeneity in ant-like colony, i.e., how the ratio of individuals (ants) obeying two different action rules affects the behavior of whole colony. For this purpose, we focus on the foraging task - searching the field for food sources, transporting the food packets to the nest. The two types of ants include what we call hard-working and lazy ants, and we perform statistical analyses to show that moderate existence of the lazy ants would boost efficient food transportation; in particular, we point out that the lazy ants play as explorer of newly emerged food sources, but also as global sensor to capture global information through their local experience, thanks to their moving-around behavior. Based on these observations, we propose a distributed estimation method of global information, i.e., to estimate the global mixture ratio by local encounter frequency with other lazy ants. Finally, we expand the estimation method to distributed control strategy of the global mixture ratio, called on-line switching strategy, where every ant dynamically alternates its obeying rules from the hard-working to the lazy and vice versa, based on its local encounter experience.