Information capacity and pattern formation in a tent map network featuring statistical periodicity

Abstract
We provide quantitative support to the observation that lattices of coupled maps are “efficient” information coding devices. It has been suggested recently that lattices of coupled maps may provide a model of information coding in the nervous system because of their ability to create structured and stimulus-dependent activity patterns which have the potential to be used for storing information. In this paper, we give an upper bound to the effective number of patterns that can be used to store information in the lattice by evaluating numerically its information capacity or information rate as a function of the coupling strength between the maps. We also estimate the time taken by the lattice to establish a limiting activity pattern.