A Study of Directional Entropy Arising from ℤ × ℤ + \mathbb{Z}\times\mathbb{Z}_{+} Semigroup Actions
Hasan Akın · 2025
Read A Study of Directional Entropy Arising from ℤ × ℤ + \mathbb{Z}\times\mathbb{Z}_{+} Semigroup Actions by Hasan Akın (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2505-10428.txt.
Introduction
Akın studies directional entropy for systems driven by two kinds of steps at once, in this 2025 chapter. Entropy here measures how fast a system creates new patterns along a chosen direction. The systems combine cellular automata, which update a grid by local rules, with a shift that slides the whole pattern.
The work gives a systematic treatment from two angles. One angle is topological, following Milnor's geometric framework for counting pattern growth. The other is measure-based, using the classic Kolmogorov-Sinai formalism. Both are developed for the combined action on a compact space of symbol sequences. The chapter ties the two views together for this class of systems.
Catalog record
| CREATOR | Hasan Akın |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 47 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2505-10428
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
A Study of Directional Entropy Arising from ℤ × ℤ + \mathbb{Z}\times\mathbb{Z}_{+} Semigroup Actions Hasan Akın Abstract This chapter investigates both topological and measure-theoretic directional entropies arising from semigroup actions generated by one-dimensional linear cellular automata (LCAs) and the shift transformation on the compact metric space ℤ m ℤ \mathbb{Z}_{m}^{\mathbb{Z}} . It aims to present a collection of formulas for computing these entropy measures. The work consolidates and builds upon previous research in the field. The study of toplological directional entropy (TDE) is carried out using Milnor’s framework, while the measure-theoretic counterpart is analyzed through the Kolmogorov–Sinai formalism. 1 Short introduction Both measure-theoretic entropy and topological entropy play fundamental roles in dynamical systems theory and ergodic theory [ 23 , 34 ] . Originally introduced by Shannon, entropy has become a crucial research topic across multiple scientific disciplines. The measure-theoretic approach to entropy, primarily developed by the Russian school, including Kolmogorov and Sinai, has led to various terminologies and methodologies for computing entropy in dynamical systems [ 24 , 32 ] . In contrast, the concept of topological entropy, first introduced by Adler et al. [ 1 ] , has been progressively extended to continuous functions on diverse spaces [ 34 ] . In 1986, Milnor introduced the concept of directional entropy [ 26 , 27 ] . Park made significant contributions to this concept [ 29 , 30 ] . Later, Kaminski and Park extended the notion by defining directional entropy for \mathbb{Z}^{2} actions on a Lebesgue space [ 22 ] . Courbage and Kaminski further explored the metric directional entropy of \mathbb{Z}^{2} -actions, particularly those arising from the natural extension of a cellular automaton (CA). This extension operates alongside the shift transformation on the space of doubly infinite sequences over a finite state space [ 15 ] . Recently, the topological entropy dimension of \mathbb{Z}^{2} -topological dynamical systems has been studied. The relationships between TDE and directional topological entropy dimension were investigated [ 25 ] . Wei et al. [ 35 ] investigate directional entropy along non-rational directions within the two-dimensional space \mathbb{R}^{2} , and they elucidate the configuration of directional Pinsker \sigma -algebra within dynamical systems preserving measures in \mathbb{Z}^{2} . In Ref. [ 13 ] , Broderick, Cyr and Kra examine the directional entropy of a dynamic system linked to a \mathbb{Z}^{2} configuration within a finite set of symbols. In a recent study by the author, detailed analyses of both measure-theoretic and topological entropy for 1D CAs on the ring \mathbb{Z}_{m} were presented [ 10 ] . In the current work, we focus on the concept of directional entropy for the \mathbb{Z}^{2} -action generated by 2D CAs. Our findings reveal that entropy values exhibit directional variations. For further insights, we suggest consulting the book chapter in [ 10 ] . In this chapter, we examine the directional measure-theoretic and topological entropies associated with the \mathbb{Z}^{2} -action, formed by the interplay between a 1D CA and a shift transformation. Building upon established research, we offer a comprehensive analysis of the topic, enriched with numerous examples and illustrative figures to facilitate comprehension. This chapter specifically investigates the directional entropy properties of \mathbb{Z}\times\mathbb{Z}_{+} -semigroup actions generated by a 1D CA F and the shift map \sigma on the full shift space \mathbb{Z}_{m}^{\mathbb{Z}} . The study employs two complementary frameworks of entropy theory. • Topological entropy (following Milnor’s axiomatic formulation): h_{\mathrm{top}}(F,\sigma)=\lim_{n\to\infty}\frac{1}{n}\log N(n,\epsilon), where N(n,\epsilon) counts (\epsilon,n) -separated orbits. • Measure-theoretic entropy (via Kolmogorov-Sinai theory): h_{\mu}(F)=\sup_{\mathscr{P}}\lim_{n\to\infty}\frac{1}{n}H_{\mu}\left(\bigvee_{k=0}^{n-1}F^{-k}\mathscr{P}\right), where \mathscr{P} denotes a measurable partition of \mathbb{Z}_{m}^{\mathbb{Z}} , and \vee is the join operation of partitions: \mathscr{A}\vee\mathscr{B}=\{A\cap B\mid A\in\mathscr{A},\ B\in\mathscr{B}\}. • LCA Entropy Formula : For linear cellular automata, h_{\mathrm{top}}(F,\sigma)=\log\lambda_{\max}(F), where \lambda_{\max}(F) is the maximal Lyapunov exponent of F . • Markov Measure Entropy : For stationary Markov measures \mu=\mu_{\pi P} with transition matrix P and stationary distribution \pi , h_{\mu}(\sigma)=-\sum_{i,j}\pi_{i}P_{ij}\log P_{ij}. 2 The MTDE of \mathbb{Z}^{2} -actions over the ring \mathbb{Z}_{m} In this section, we investigate the MTDEs of \mathbb{Z}^{2} -actions generated by 1D LCA and the shift map. Let us give some necessary definitions. Definition 1 [ 20 ] Let T_{f[l,r]}:\mathbb{Z}_{m}^{\mathbb{Z}}\to\mathbb{Z}_{m}^{\mathbb{Z}} be a 1D- LCA associated to the local rule f with minimal memory Mem[l,r] . The local rule f and the 1D- LCA T_{f[l,r]} are called: (1) leftmost permutive if l<0 and f is permutative in x_{l} ; (2) rightmost permutive if r>0 and f is permutative in x_{r} ; (3) bilaterally permutative (bipermutative) if it is both leftmost and rightmost permutative. The investigation of invariant measures, a cornerstone of ergodic theory, has been rigorously explored through diverse methodological lenses [ 31 ] . Notably, Chang and Chen [ 14 ] established explicit formulas for both measure-theoretic and topological entropy of weakly permutive CAs under invariant measures on the configuration space \mathbb{Z}_{m}^{\mathbb{Z}} . A critical property of 1D CAs F_{f[l,r]} is their commutativity with the shift map \sigma [ 21 ] , necessitating the characterization of invariant measures for the semigroup action generated by F_{f[l,r]} and \sigma . Let \mathcal{M}(\mathbb{Z}_{m}^{\mathbb{Z}}) denote the space of probability measures on \mathbb{Z}_{m}^{\mathbb{Z}} defined over the \sigma -algebra \mathcal{B} . A measure \mu\in\mathcal{M}(\mathbb{Z}_{m}^{\mathbb{Z}}) is termed \sigma -invariant (respectively F_{f[l,r]} -invariant) if \sigma\mu=\mu (respectively F_{f[l,r]}\mu=\mu ). Consequently, \mu is (F_{f[l,r]},\sigma) -invariant if it is invariant under both transformations [ 31 ] . For CAs governed by local rules dependent on finite neighborhoods, F_{f[l,r]} exhibits invariance under spatial translation \sigma . This permits defining the topological entropy of F_{f[l,r]} along direction (p,q)\in\mathbb{Z}\times\mathbb{N} as the entropy of F_{f[l,r]}^{q}\circ\sigma^{p} [ 12 ] . Consider a normalized ergodic measure \mu invariant under the \mathbb{Z}^{2} -action. Let h_{m,n}=h(\sigma^{m}F^{n},\mu) denote the measure-theoretic entropy of the transformation \sigma^{m}F^{n} , where F=F_{f[l,r]} is a 1D linear CA (LCA) over \mathbb{Z}_{m} . This subsection focuses on Bernoulli and Markov measures within the measurable space (\mathbb{Z}_{m}^{\mathbb{Z}},\mathcal{B}) . Define the partition \xi=\{\!_{0}[0],\!_{0}[1],\ldots,\!_{0}[p-1]\} of \mathbb{F}_{p}^{\mathbb{Z}} , and let \xi_{m,n}=\Phi^{(m,n)}\xi=\sigma^{m}F_{f[l,r]}^{n}(\xi), where \Phi^{(m,n)} represents the joint action of \sigma^{m} and F_{f[l,r]}^{n} . To calculate the directional entropy, Sinai [ 33 ] introduced standard notation and fundamental concepts from the theory of measurable partitions and measure-theoretic entropy. The segment on the plane connecting the points (a,0) and (a+\omega^{-l},1) is denoted by I=I(a,\omega) , while \Gamma(a,\omega) represents the half-line given by y=\omega(x-a) for y\leq 1 . Clearly, I(a,\omega) is contained within \Gamma(a,\omega) . Throughout, we will assume \omega>0 . Figure 1: The segment I(a,\omega) on the plane joining the points (a,0) and (a+\omega^{-l},1) . Definition 2 Let (X,\mathcal{A},\mu) be a measure space. Let \alpha and \beta be two partitions of X . The quantity H_{\mu}(\alpha|\beta)=-\sum\limits_{B\in\beta}\mu(B)\sum\limits_{A\in\alpha}\mu(A|B)\log\mu(A|B) is called the conditional entropy of the partition \alpha given \beta . To prove the theorems, Sinai [ 33 ] deals with the following conditional entropies: \begin{array}[]{l}\mathcal{H}_{r}(I)=H\left(\underset{m\geq a+\omega^{-1}}{\bigvee}\xi_{m,1}|\underset{n=0}{\overset{\infty}{\bigvee}}\underset{m\geq a+\omega^{-1}n}{\bigvee}\xi_{m,-n}\right)\\ \mathcal{H}_{l}(I)=H\left(\underset{m\leq a+\omega^{-1}}{\bigvee}\xi_{m,1}|\underset{n=0}{\overset{\infty}{\bigvee}}\underset{m\leq a+\omega^{-1}n}{\bigvee}\xi_{m,-n}\right)\\ \mathcal{H}(I)=\mathcal{H}_{l}(I)+\mathcal{H}_{r}(I).\end{array} Let us consider a transformation Q in the space of segments I(a,\omega) , where Q(I(a,\omega))=I(a^{\prime},to) , a^{\prime}=a+\omega^{-1} . Our first result is the following theorem. Theorem 2.1 ( 33 , Theorem 1) Let p>0,q>0 have no common factor. Then h_{p,q}=\sum_{i=0}^{p-1}\mathcal{H}(Q^{i}(I))=p\int_{0}^{1}\mathcal{H}(I)da for any interval I=I(a,-q/p) . We recall the definitions and properties of the MTDE for \mathbb{Z}^{2} -actions, using elementary entropy properties (see [ 22 , 29 , 30 , 32 ] for details). Let (X,\mathcal{B},\mu) be a Lebesgue probability space, and let \mathcal{Z} be the