Filters defined by random contexts versus polarization in networks of evolutionary processors

Abstract

The filtering process of Networks of Evolutionary Processors (NEP for short) is regulated by random-context conditions or polarization. Both models, the NEPs with random-context filters and the NEPs with polarization, are computationally complete, hence they are equivalent from the computational power point of view. The computationally completeness results have been obtained by simulations of other universal models like Turing machines and 2-tag systems. We start here the study of a direct simulation between the two models. The main result of this paper is that each NEP with random-context filters can be simulated by NEPs with polarization. Furthermore, the simulation preserves the time complexity and the size of the simulating network depends linearly on the size of the simulated network. The converse simulation remains to be investigated.