Entropy Analysis of Noise Contaminated Sequences

Abstract

The entropy analysis of symbol sequences created by symbolic dynamics reveals temporal correlations present in the stationary dynamics. We investigate how noise, introduced in different ways, affects temporal correlations found in the logistic map at the period doubling accumulation point. The deterministic system under consideration yields a highly structured spectrum of conditional entropies. We will discuss how additive noise, parametric noise, a fluctuating cell boundary and a noisy transmission channel modify the unperturbed entropy spectrum.