Locking-time and Information Capacity in CML with Statistical Periodicity

Abstract

In this work we address the statistical periodicity phenomenon on a coupled map lattice. The study was done based on the asymptotic binary patterns. The pattern multiplicity gives us the lattice information capacity, while the entropy rate allows us to calculate the locking-time. Our results suggest that the lattice has low locking-time and high capacity information when the coupling is weak. This is the condition for the system to reproduce a kind of behavior observed in neural networks.