Oscillations and Chaos in Ant Societies

A neural network-like model of collective short-time oscillations in ant colonies is presented. Such behaviour has been recently observed in some experimental situations. Each individual is here considered as a cellular automaton able both to move into a given available space and to interact with other (nearest) automata. As a consequence of non-linear interactions, the observed oscillations are an emergent property of the colony as a whole. Time series and Fourier spectrum are in agreement with real data. The internal dynamics of each individual is modelled either by random process or deterministic chaos.