Entropy for symbolic dynamics with overlapping alphabets (Q300352)

The authors consider shift spaces in which elements of the alphabet may overlap non-transitively. While most of the results known in the field of symbolic dynamics arises when the partition of the space consists in regions not overlapping each others (except on their boundaries), the authors weaken this hypothesis. This allows them to use this kind of symbolic dynamical systems to model a large variety of (more complicated) dynamical systems. In the classical case, that is when the elements of the shift do not overlap, it is well known how to extract useful dynamical information from such a system. Indeed it is enough to compute the growth rate of the number of words of length \(n\) in the shift space to obtain a lower bound for the entropy. In the case described in the paper an orbit may have multiple itineraries, thus these classical methods cannot be used. In this paper a notion of entropy is defined for the case when the neighbourhoods of the partition intersect. This new entropy gives a lower bound for the topological entropy. When the underlying shift is of finite type, the entropy here defined is equal to a limit of standard (non-overlapping) entropies. Moreover, it is shown that when such a shift space with overlaps arises as a model for a discrete dynamical system with a finite set of overlapping neighbourhoods the entropy gives a lower bound for the topological entropy of the dynamical system.