Article Outline
Glossary
Definition of the Subject
Introduction
Invariant Measures for CA
Limit Measures and Other Asymptotics
Measurable Dynamics
Entropy
Future Directions and Open Problems
Acknowledgments
Bibliography
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsAbbreviations
- Configuration space and the shift :
-
Let \( { {\mathbb{M}} } \) be a finitely generated group or monoid (usually abelian). Typically, \( { {\mathbb{M}}=\mathbb{N}:=\{0,1,2,\ldots\} } \) or \( {\mathbb{M}}=\mathbb{Z}:=\{\ldots, -1,0,1,2,\ldots\} \), or \( { {\mathbb{M}}=\mathbb{N}^E } \), \( { \mathbb{Z} } \) D, or \( { \mathbb{Z}^D\times\mathbb{N}^E } \) for some \( { D,E\in\mathbb{N} } \). In some applications, \( { {\mathbb{M}} } \) could be nonabelian (although usually amenable), but to avoid notational complexity we will generally assume \( { {\mathbb{M}} } \) is abelian and additive, with operation ‘+’.
Let \( { {\mathcal{A}} } \) be a finite set of symbols (called an alphabet ). Let \( { \mathcal{A}^{\mathbb{M}} } \) denote the set of all functions \( { {\mathbf{a}}\colon{\mathbb{M}}{\longrightarrow}{\mathcal{A}} } \), which we regard as \( { {\mathbb{M}} } \)?indexed configurations of elements in \( { {\mathcal{A}} } \). We write such a configuration as \( { {\mathbf{a}}=[a_{\textsf{m}}]_{{\textsf{m}}\in{\mathbb{M}}} } \), where \( { a_{\textsf{m}}\in{\mathcal{A}} } \) for all \( { {\textsf{m}}\in{\mathbb{M}} } \), and refer to \( { \mathcal{A}^{\mathbb{M}} } \) as configuration space .
Treat \( { {\mathcal{A}} } \) as a discrete topological space; then \( { {\mathcal{A}} } \) is compact (because it is finite), so \( { \mathcal{A}^{\mathbb{M}} } \) is compact in the Tychonoff product topology. In fact, \( { \mathcal{A}^{\mathbb{M}} } \) is a Cantor space : it is compact, perfect, totally disconnected, and metrizable. For example, if \( { {\mathbb{M}}=\mathbb{Z}^D } \), then the standard metric on \( { \mathcal{A}^{\mathbb{Z}^D} } \) is defined \( { d({\mathbf{a}},{\mathbf{b}})=2^{-\Delta({\mathbf{a}},{\mathbf{b}})} } \), where \( { \Delta({\mathbf{a}},{\mathbf{b}}):=\min{\{|{\textsf{z}}|; a_{\textsf{z}}\neq b_{\textsf{z}}\}} } \).
Any \( { {\textsf{v}}\in{\mathbb{M}} } \), determines a continuous shift map \( { \sigma^v\colon \mathcal{A}^{\mathbb{M}}{\longrightarrow}\mathcal{A}^{\mathbb{M}} } \) defined by \( { \sigma^v (\mathbf{a})_{\textsf{m}} = a_{{\textsf{m}}+{\textsf{v}}} } \) for all \( { {\mathbf{a}}\in\mathcal{A}^{\mathbb{M}} } \) and \( { {\textsf{m}}\in{\mathbb{M}} } \). The set \( { \{\sigma^v\}_{{\textsf{v}}\in{\mathbb{M}}} } \) is then a continuous \( { {\mathbb{M}} } \)?action on \( { \mathcal{A}^{\mathbb{M}} } \), which we denote simply by “σ”.
If \( { {\mathbf{a}}\in\mathcal{A}^{\mathbb{M}} } \) and \( { {\mathbb{U}}\subset{\mathbb{M}} } \), then we define \( { {\mathbf{a}}_{\mathbb{U}}\in{\mathcal{A}}^{\mathbb{U}} } \) by \( { {\mathbf{a}}_{\mathbb{U}}:=[a_{\textsf{u}}]_{{\textsf{u}}\in{\mathbb{U}}} } \). If \( { {\textsf{m}}\in{\mathbb{M}} } \), then strictly speaking, \( { {\mathbf{a}}_{{\textsf{m}}+{\mathbb{U}}}\in{\mathcal{A}}^{{\textsf{m}}+{\mathbb{U}}} } \); however, it will often be convenient to ‘abuse notation’ and treat \( { {\mathbf{a}}_{{\textsf{m}}+{\mathbb{U}}} } \) as an element of \( { {\mathcal{A}}^{{\mathbb{U}}} } \) in the obvious way.
- Cellular automata :
-
Let \( { \mathbb{H}\subset{\mathbb{M}} } \) be some finite subset, and let \( { \phi\colon{\mathcal{A}}^{\mathbb{H}}{\longrightarrow}{\mathcal{A}} } \) be a function (called a local rule ). The cellular automaton (CA) determined by ? is the function \( { \Phi\colon\mathcal{A}^{\mathbb{M}}{\longrightarrow}\mathcal{A}^{\mathbb{M}} } \) defined by \( { \Phi({\mathbf{a}})_{{\textsf{m}}} = \phi({\mathbf{a}}_{{\textsf{m}}+\mathbb{H}}) } \) for all \( { {\mathbf{a}}\in\mathcal{A}^{\mathbb{M}} } \) and \( { {\textsf{m}}\in{\mathbb{M}} } \). Curtis, Hedlund and Lyndon showed that cellular automata are exactly the continuous transformations of \( { \mathcal{A}^{\mathbb{M}} } \) which commute with all shifts (see Theorem 3.4 in [58]). We refer to \( { \mathbb{H} } \) as the neighborhood of Φ. For example, if \( { {\mathbb{M}}=\mathbb{Z} } \), then typically \( { \mathbb{H}:=[-\ell \ldots r]:=\{-\ell,1-\ell,\ldots,r-1,r\} } \) for some left radius \( { \ell \geq 0 } \) and right radius \( { r \geq 0 } \). If \( { -\ell \geq 0 } \), then ? can either define CA on \( { \mathcal{A}^{\mathbb{N}} } \) or define a one-sided CA on \( { \mathcal{A}^{\mathbb{Z}} } \). If \( { {\mathbb{M}}=\mathbb{Z}^D } \), then typically \( { \mathbb{H}\subseteq[-R \ldots R]^D } \), for some radius \( { R\geq 0 } \). Normally we assume that \( { \ell } \), r, and R are chosen to be minimal. Several specific classes of CA will be important to us:
- Linear CA :
-
Let \( { ({\mathcal{A}},+) } \) be a finite abelian group (e.?g. \( { {\mathcal{A}}=\mathbb{Z}_{/p} } \), where \( { p\in\mathbb{N} } \); usually p is prime). Then Φ is a linear CA (LCA) if the local rule ? has the form
$$ \phi({\mathbf{a}}_\mathbb{H}) := \sum_{\textsf{h}\in\mathbb{H}} \varphi_{\textsf{h}}(a_{\textsf{h}})\:,\quad \forall \mathbf{a}_\mathbb{H} \in \mathcal{A}^{\mathbb{H}}\:, $$(1)where \( { \varphi_{\textsf{h}}\colon\mathcal{A}{\longrightarrow}\mathcal{A} } \) is an endomorphism of \( { (\mathcal{A},+) } \), for each \( { \textsf{h}\in\mathbb{H} } \). We say that Φ has scalar coefficients if, for each \( { \textsf{h}\in\mathbb{H} } \), there is some scalar \( { c_{\textsf{h}}\in\mathbb{Z} } \), so that \( { \varphi_{\textsf{h}}(a_{\textsf{h}}) := c_{\textsf{h}} \cdot a_{\textsf{h}} } \); then \( { \phi({\mathbf{a}}_\mathbb{H}) := \sum_{\textsf{h}\in\mathbb{H}} c_{\textsf{h}} a_{\textsf{h}} } \). For example, if \( \mathcal{A}=(\mathbb{Z}_{/p},+) \), then all endomorphisms are scalar multiplications, so all LCA have scalar coefficients.
If \( { c_{\textsf{h}}=1 } \) for all \( { \textsf{h}\in\mathbb{H} } \), then Φ has local rule \( { \phi(\mathbf{a}_\mathbb{H}) := \sum_{\textsf{h}\in\mathbb{H}} a_{\textsf{h}} } \); in this case, Φ is called an additive cellular automaton ; see Additive Cellular Automata.
- Affine CA :
-
If \( { ({\mathcal{A}},+) } \) is a finite abelian group, then an affine CA is one with a local rule \( \phi({\mathbf{a}}_\mathbb{H}) := c+ \sum_{\textsf{h}\in\mathbb{H}} \varphi_{\textsf{h}}(a_{\textsf{h}}) \), where c is some constant and where \( { \varphi_{\textsf{h}}\colon{\mathcal{A}}{\longrightarrow}{\mathcal{A}} } \) are endomorphisms of \( { ({\mathcal{A}},+) } \). Thus, Φ is an LCA if \( { c=0 } \).
- Permutative CA :
-
Suppose \( \Phi\colon\mathcal{A}^{\mathbb{Z}}{\longrightarrow}\mathcal{A}^{\mathbb{Z}} \) has local rule \( \phi\colon{\mathcal{ A}}^{[-\ell \ldots r]}{\longrightarrow}{\mathcal{A}} \). Fix \( {\mathbf{b}}=[b_{1-\ell},\ldots, b_{r-1}, b_r]\in{\mathcal{A}}^{(-\ell \ldots r]} \). For any \( { a\in{\mathcal{A}} } \), define \( [a\;{\mathbf{b}}]:=[a,b_{1-\ell},\ldots,b_{r-1},b_r]\in {\mathcal{A}}^{[-\ell \ldots r]} \). We then define the function \( \phi_{\mathbf{b}}\colon{\mathcal{A}}{\longrightarrow}{\mathcal{A}} \) by \( \phi_{\mathbf{b}}(a) := \phi([a\;{\mathbf{b}}]) \). We say that Φ is left‐permutative if \( \phi_{\mathbf{b}}\colon{\mathcal{A}}{\longrightarrow}{\mathcal{A}} \) is a permutation (i.?e. a bijection) for all \( { \mathbf{b}\in\mathcal{A}^{(-\ell \ldots r]} } \).
Likewise, given \( { {\mathbf{b}}=[b_{-\ell},\ldots,b_{r-1}]\in{\mathcal{A}}^{[-\ell \ldots r)} } \) and \( { c\in{\mathcal{A}} } \), define \( [{\mathbf{b}}\;c]:=[b_{-\ell},b_{1-\ell}\ldots,b_{r-1},c]\in{\mathcal{A}}^{[-\ell \ldots r]} \), and define \( _{\mathbf{b}}\phi\colon{\mathcal{A}}{\longrightarrow}{\mathcal{A}} \) by \( _{\mathbf{b}}\phi(c) := \phi([{\mathbf{b}} c]) \); then Φ is right‐permutative if \( _{\mathbf{b}}\phi\colon {\mathcal{A}}{\longrightarrow} {\mathcal{A}} \) is a permutation for all \( {\mathbf{b}}\in{\mathcal{A}}^{[-\ell \ldots r)} \). We say Φ is bipermutative if it is both left- and right‐permutative. More generally, if \( { {\mathbb{M}} } \) is any monoid, \( { \mathbb{H}\subset{\mathbb{M}} } \) is any neighborhood, and \( { \textsf{h}\in\mathbb{H} } \) is any fixed coordinate, then we define h?permutativity for a CA on \( { \mathcal{A}^{\mathbb{M}} } \) in the obvious fashion.
For example, suppose \( { ({\mathcal{A}},+) } \) is an abelian group and Φ is an affine CA on \( { \mathcal{A}^{\mathbb{Z}} } \) with local rule \( \phi({\mathbf{a}}_\mathbb{H}) = c + \sum_{h=-\ell}^r \varphi_h(a_h) \). Then Φ is left‐permutative iff \( { \varphi_{-\ell} } \) is an automorphism, and right‐permutative iff \( { \varphi_r } \) is an automorphism. If \( { {\mathcal{A}}=\mathbb{Z}_{/p} } \), and p is prime, then every nontrivial endomorphism is an automorphism(because it is multiplication by a nonzero element of \( { \mathbb{Z}_{/p} } \), which is a field), so in this case, every affine CA is permutative in every coordinate of its neighborhood (and in particular, bipermutative). If \( { {\mathcal{A}}\neq\mathbb{Z}_{/p} } \), however, then not all affine CA are permutative.
Permutative CA were introduced by Hedlund [58], §6, and are sometimes called permutive CA. Right permutative CA on \( \smash{ \mathcal{A}^{\mathbb{N}} } \) are also called toggle automata . For more information, see Sect. 7 of Topological Dynamics of Cellular Automata.
- Linear CA :
-
Let \( { ({\mathcal{A}},+) } \) be a finite abelian group (e.?g. \( { {\mathcal{A}}=\mathbb{Z}_{/p} } \), where \( { p\in\mathbb{N} } \); usually p is prime). Then Φ is a linear CA (LCA) if the local rule ? has the form
$$ \phi({\mathbf{a}}_\mathbb{H}) := \sum_{\textsf{h}\in\mathbb{H}} \varphi_{\textsf{h}}(a_{\textsf{h}})\:,\quad \forall \mathbf{a}_\mathbb{H} \in \mathcal{A}^{\mathbb{H}}\:, $$(1)where \( { \varphi_{\textsf{h}}\colon\mathcal{A}{\longrightarrow}\mathcal{A} } \) is an endomorphism of \( { (\mathcal{A},+) } \), for each \( { \textsf{h}\in\mathbb{H} } \). We say that Φ has scalar coefficients if, for each \( { \textsf{h}\in\mathbb{H} } \), there is some scalar \( { c_{\textsf{h}}\in\mathbb{Z} } \), so that \( { \varphi_{\textsf{h}}(a_{\textsf{h}}) := c_{\textsf{h}} \cdot a_{\textsf{h}} } \); then \( { \phi({\mathbf{a}}_\mathbb{H}) := \sum_{\textsf{h}\in\mathbb{H}} c_{\textsf{h}} a_{\textsf{h}} } \). For example, if \( \mathcal{A}=(\mathbb{Z}_{/p},+) \), then all endomorphisms are scalar multiplications, so all LCA have scalar coefficients.
If \( { c_{\textsf{h}}=1 } \) for all \( { \textsf{h}\in\mathbb{H} } \), then Φ has local rule \( { \phi(\mathbf{a}_\mathbb{H}) := \sum_{\textsf{h}\in\mathbb{H}} a_{\textsf{h}} } \); in this case, Φ is called an additive cellular automaton ; see Additive Cellular Automata.
- Affine CA :
-
If \( { ({\mathcal{A}},+) } \) is a finite abelian group, then an affine CA is one with a local rule \( \phi({\mathbf{a}}_\mathbb{H}) := c+ \sum_{\textsf{h}\in\mathbb{H}} \varphi_{\textsf{h}}(a_{\textsf{h}}) \), where c is some constant and where \( { \varphi_{\textsf{h}}\colon{\mathcal{A}}{\longrightarrow}{\mathcal{A}} } \) are endomorphisms of \( { ({\mathcal{A}},+) } \). Thus, Φ is an LCA if \( { c=0 } \).
- Permutative CA :
-
Suppose \( \Phi\colon\mathcal{A}^{\mathbb{Z}}{\longrightarrow}\mathcal{A}^{\mathbb{Z}} \) has local rule \( \phi\colon{\mathcal{ A}}^{[-\ell \ldots r]}{\longrightarrow}{\mathcal{A}} \). Fix \( {\mathbf{b}}=[b_{1-\ell},\ldots, b_{r-1}, b_r]\in{\mathcal{A}}^{(-\ell \ldots r]} \). For any \( { a\in{\mathcal{A}} } \), define \( [a\;{\mathbf{b}}]:=[a,b_{1-\ell},\ldots,b_{r-1},b_r]\in {\mathcal{A}}^{[-\ell \ldots r]} \). We then define the function \( \phi_{\mathbf{b}}\colon{\mathcal{A}}{\longrightarrow}{\mathcal{A}} \) by \( \phi_{\mathbf{b}}(a) := \phi([a\;{\mathbf{b}}]) \). We say that Φ is left‐permutative if \( \phi_{\mathbf{b}}\colon{\mathcal{A}}{\longrightarrow}{\mathcal{A}} \) is a permutation (i.?e. a bijection) for all \( { \mathbf{b}\in\mathcal{A}^{(-\ell \ldots r]} } \).
Likewise, given \( { {\mathbf{b}}=[b_{-\ell},\ldots,b_{r-1}]\in{\mathcal{A}}^{[-\ell \ldots r)} } \) and \( { c\in{\mathcal{A}} } \), define \( [{\mathbf{b}}\;c]:=[b_{-\ell},b_{1-\ell}\ldots,b_{r-1},c]\in{\mathcal{A}}^{[-\ell \ldots r]} \), and define \( _{\mathbf{b}}\phi\colon{\mathcal{A}}{\longrightarrow}{\mathcal{A}} \) by \( _{\mathbf{b}}\phi(c) := \phi([{\mathbf{b}} c]) \); then Φ is right‐permutative if \( _{\mathbf{b}}\phi\colon {\mathcal{A}}{\longrightarrow} {\mathcal{A}} \) is a permutation for all \( {\mathbf{b}}\in{\mathcal{A}}^{[-\ell \ldots r)} \). We say Φ is bipermutative if it is both left- and right‐permutative. More generally, if \( { {\mathbb{M}} } \) is any monoid, \( { \mathbb{H}\subset{\mathbb{M}} } \) is any neighborhood, and \( { \textsf{h}\in\mathbb{H} } \) is any fixed coordinate, then we define h?permutativity for a CA on \( { \mathcal{A}^{\mathbb{M}} } \) in the obvious fashion.
For example, suppose \( { ({\mathcal{A}},+) } \) is an abelian group and Φ is an affine CA on \( { \mathcal{A}^{\mathbb{Z}} } \) with local rule \( \phi({\mathbf{a}}_\mathbb{H}) = c + \sum_{h=-\ell}^r \varphi_h(a_h) \). Then Φ is left‐permutative iff \( { \varphi_{-\ell} } \) is an automorphism, and right‐permutative iff \( { \varphi_r } \) is an automorphism. If \( { {\mathcal{A}}=\mathbb{Z}_{/p} } \), and p is prime, then every nontrivial endomorphism is an automorphism(because it is multiplication by a nonzero element of \( { \mathbb{Z}_{/p} } \), which is a field), so in this case, every affine CA is permutative in every coordinate of its neighborhood (and in particular, bipermutative). If \( { {\mathcal{A}}\neq\mathbb{Z}_{/p} } \), however, then not all affine CA are permutative.
Permutative CA were introduced by Hedlund [58], §6, and are sometimes called permutive CA. Right permutative CA on \( \smash{ \mathcal{A}^{\mathbb{N}} } \) are also called toggle automata . For more information, see Sect. 7 of Topological Dynamics of Cellular Automata.
- Subshifts :
-
A subshift is a closed, σ?invariant subset \( \mathbf{X}\subset\mathcal{A}^{\mathbb{M}} \). For any \( \smash{ \mathbb{U}\subset{\mathbb{M}} } \), let \( \smash{\mathbf{X}}_{\mathbb{U}} := {\{{\mathbf{x}}_{\mathbb{U}}; {\mathbf{x}}\in{\mathbf{X}}\} }\subset {\mathcal{A}}^{\mathbb{U}} \). We say \( \smash{ {\mathbf{X}} } \) is a subshift of finite type (SFT) if there is some finite \( \smash{ {\mathbb{U}}\subset{\mathbb{M}} } \) such that \( \smash{ {\mathbf{X}} } \) is entirely described by \( \smash{ {\mathbf{X}}_{\mathbb{U}} } \), in the sense that \( \smash{\mathbf{X}}={\{{\mathbf{x}}\in\mathcal{A}^{\mathbb{M}} ; {\mathbf{x}}_{{\mathbb{U}}+{\textsf{ m}}}\in{\mathbf{X}}_{\mathbb{U}}, \forall{\textsf{m}}\in{\mathbb{M}}\}} \).
In particular, if \( \smash{ {\mathbb{M}}=\mathbb{Z} } \), then a (two-sided) Markov subshift is an SFT \( \smash{ {\mathbf{X}}\subset\mathcal{A}^{\mathbb{Z}} } \) determined by a set \( \smash{ {\mathbf{X}}_{\{0,1\}}\subset{\mathcal{ A}}^{\{0,1\}} } \) of admissible transitions ; equivalently, \( \smash{ {\mathbf{X}} } \) is the set of all bi-infinite directed paths in a digraph whose vertices are the elements of \( \smash{ {\mathcal{A}} } \), with an edge \( \smash{ a\leadsto b } \) iff \( \smash{ (a,b)\in{\mathbf{X}}_{\{0,1\}} } \). If \( \smash{ {\mathbb{M}}=\mathbb{N} } \), then a one-sided Markov subshift is a subshift of \( \smash{ \mathcal{A}^{\mathbb{N}} } \) defined in the same way.
If \( \smash{ D\geq 2 } \), then an SFT in \( \smash{ \mathcal{A}^{\mathbb{Z}^D} } \) can be thought of as the set of admissible ‘tilings’ of \( \smash{ \mathbb{R} } \) D by Wang tiles corresponding to the elements of \( \smash{ {\mathbf{X}}_{\mathbb{U}} } \). (Wang tiles are unit squares (or (hyper)cubes) with various ‘notches’ cut into their edges (or (hyper)faces) so that they can only be juxtaposed in certain ways.)
A subshift \( \smash{ {\mathbf{X}}\subseteq\mathcal{A}^{\mathbb{Z}^D} } \) is strongly irreducible (or topologically mixing ) if there is some \( \smash{ R\in\mathbb{N} } \) such that, for any disjoint finite subsets \( \smash{ {\mathbb{V}},{\mathbb{U}}\subset\mathbb{Z}^D } \) separated by a distance of at least R, and for any \( \smash{ {\mathbf{u}}\in{\mathbf{X}}_{\mathbb{U}} } \) and \( \smash{ {\mathbf{v}}\in{\mathbf{ X}}_{\mathbb{V}} } \), there is some \( \smash{ {\mathbf{x}}\in{\mathbf{X}} } \) such that \( \smash{ {\mathbf{ x}}_{\mathbb{U}}={\mathbf{u}} } \) and \( \smash{ {\mathbf{x}}_{\mathbb{V}}={\mathbf{v}} } \).
- Measures :
-
For any finite subset \( \smash{ {\mathbb{U}}\subset{\mathbb{M}} } \), and any \( \smash{ {\mathbf{b}}\in{\mathcal{A}}^{\mathbb{U}} } \), let \( \smash{ \langle{\mathbf{b}}\rangle := {\{{\mathbf{a}}\in\mathcal{A}^{\mathbb{M}} ; {\mathbf{a}}_{\mathbb{U}}:={\mathbf{b}}\} } } \) be the cylinder set determined by \( \smash{ {\mathbf{b}} } \). Let \( \smash{ {\mathfrak{B}} } \) be the sigma-algebra on \( \smash{ \mathcal{A}^{\mathbb{M}} } \) generated by all cylinder sets. A (probability) measure μ on \( \smash{ \mathcal{A}^{\mathbb{M}} } \) is a countably additive function \( \smash{ \mu\colon{\mathfrak{ B}}{\longrightarrow}[0,1] } \) such that \( \smash{ \mu[\mathcal{A}^{\mathbb{M}}]=1 } \). A measure on \( \smash{ \mathcal{A}^{\mathbb{M}} } \) is entirely determined by its values on cylinder sets. We will be mainly concerned with the following classes of measures:
- Bernoulli measure :
-
Let \( { \beta_0 } \) be a probability measure on \( { {\mathcal{A}} } \). The Bernoulli measure induced by \( { \beta_0 } \) is the measure β on \( { \mathcal{A}^{\mathbb{M}} } \) such that, for any finite subset \( { {\mathbb{U}}\subset{\mathbb{M}} } \), and any \( { {\mathbf{a}}\in{\mathcal{A}}^{\mathbb{U}} } \), if \( { U:=|{\mathbb{U}}| } \), then \( { \beta[\langle{\mathbf{a}}\rangle] = \prod_{\textsf{h}\in\mathbb{H}} \beta_0(a_{\textsf{h}}) } \).
- Invariant measure :
-
Let μ be a measure on \( { \mathcal{A}^{\mathbb{M}} } \), and let \( { \Phi\colon\mathcal{A}^{\mathbb{M}}{\longrightarrow}\mathcal{A}^{\mathbb{M}} } \) be a cellular automaton. The measure \( { \Phi\mu } \) is defined by \( { \Phi\mu ({\mathbf{B}}) = \mu(\Phi^{-1}({\mathbf{B}})) } \), for any \( { {\mathbf{B}}\in{\mathfrak{B}} } \). We say that μ is Φ?invariant (or that Φ is μ?preserving ) if \( { \Phi\mu = \mu } \).
- Uniform measure :
-
Let \( { A:=|{\mathcal{A}}| } \). The uniform measure η on \( { \mathcal{A}^{\mathbb{M}} } \) is the Bernoulli measure such that, for any finite subset \( { {\mathbb{U}}\subset{\mathbb{M}} } \), and any \( { {\mathbf{b}}\in{\mathcal{A}}^{\mathbb{U}} } \), if \( { U:=|{\mathbb{U}}| } \), then \( { \mu[\langle{\mathbf{b}}\rangle] = 1/A^U } \).
The support of a measure μ is the smallest closed subset \( { \mathbf{X}\subset\mathcal{A}^{\mathbb{M}} } \) such that \( { \mu[{\mathbf{X}}]=1 } \); we denote this by supp(μ). We say μ has full support if \( \textsf{supp}(\mu)=\mathcal{A}^{\mathbb{M}} \) – equivalently, \( { \mu[{\mathbf{C}}] > 0 } \) for every cylinder subset \( { {\mathbf{C}}\subset\mathcal{A}^{\mathbb{M}} } \).
- Bernoulli measure :
-
Let \( { \beta_0 } \) be a probability measure on \( { {\mathcal{A}} } \). The Bernoulli measure induced by \( { \beta_0 } \) is the measure β on \( { \mathcal{A}^{\mathbb{M}} } \) such that, for any finite subset \( { {\mathbb{U}}\subset{\mathbb{M}} } \), and any \( { {\mathbf{a}}\in{\mathcal{A}}^{\mathbb{U}} } \), if \( { U:=|{\mathbb{U}}| } \), then \( { \beta[\langle{\mathbf{a}}\rangle] = \prod_{\textsf{h}\in\mathbb{H}} \beta_0(a_{\textsf{h}}) } \).
- Invariant measure :
-
Let μ be a measure on \( { \mathcal{A}^{\mathbb{M}} } \), and let \( { \Phi\colon\mathcal{A}^{\mathbb{M}}{\longrightarrow}\mathcal{A}^{\mathbb{M}} } \) be a cellular automaton. The measure \( { \Phi\mu } \) is defined by \( { \Phi\mu ({\mathbf{B}}) = \mu(\Phi^{-1}({\mathbf{B}})) } \), for any \( { {\mathbf{B}}\in{\mathfrak{B}} } \). We say that μ is Φ?invariant (or that Φ is μ?preserving ) if \( { \Phi\mu = \mu } \).
- Uniform measure :
-
Let \( { A:=|{\mathcal{A}}| } \). The uniform measure η on \( { \mathcal{A}^{\mathbb{M}} } \) is the Bernoulli measure such that, for any finite subset \( { {\mathbb{U}}\subset{\mathbb{M}} } \), and any \( { {\mathbf{b}}\in{\mathcal{A}}^{\mathbb{U}} } \), if \( { U:=|{\mathbb{U}}| } \), then \( { \mu[\langle{\mathbf{b}}\rangle] = 1/A^U } \).
- Notation :
-
Let \( { {\textsf{C\!A}(\mathcal{A}^{\mathbb{M}})} } \) denote the set of all cellular automata on \( { \mathcal{A}^{\mathbb{M}} } \). If \( { {\mathbf{X}}\subset\mathcal{A}^{\mathbb{M}} } \), then let \( { {\textsf{C\!A}(\mathbf{X})} } \) be the subset of all \( { \Phi\in{\textsf{C\!A}(\mathcal{A}^{\mathbb{M}})} } \) such that \( { \Phi({\mathbf{X}})\subseteq{\mathbf{X}} } \). Let \( { \mathfrak{M\scriptscriptstyle{\!e\!a\!s}}(\mathcal{A}^{\mathbb{M}}) } \) be the set of all probability measures on \( { \mathcal{A}^{\mathbb{M}} } \), and let \( { \mathfrak{M\scriptscriptstyle{\!e\!a\!s}}(\mathcal{A}^{\mathbb{M}};\Phi) } \) be the subset of Φ?invariant measures. If \( { {\mathbf{X}}\subset\mathcal{A}^{\mathbb{M}} } \), then let \( { \mathfrak{M\scriptscriptstyle{\!e\!a\!s}}({\mathbf{X}}) } \) be the set of probability measures μ with \( { {\textsf{supp}}(\mu)\subseteq{\mathbf{X}} } \), and define \( { \mathfrak{M\scriptscriptstyle{\!e\!a\!s}}({\mathbf{X}};\Phi) } \) in the obvious way.
- Font conventions :
-
Upper case calligraphic letters (\( { {\mathcal{A}},{\mathcal{B}},{\mathcal{C}},\ldots } \)) denote finite alphabets or groups. Upper-case bold letters (\( { {\mathbf{A}},{\mathbf{B}},{\mathbf{C}},\ldots } \)) denote subsets of \( { \mathcal{A}^{\mathbb{M}} } \) (e.?g. subshifts), lowercase bold-faced letters (\( { {\mathbf{a}},{\mathbf{b}},{\mathbf{c}},\ldots } \)) denote elements of \( { \mathcal{A}^{\mathbb{M}} } \), and Roman letters (\( { a,b,c,\ldots } \)) are elements of \( { {\mathcal{A}} } \) or ordinary numbers. Lower-case sans-serif (\( { \ldots,{\textsf{m}},{\textsf{n}},{\textsf{p}} } \)) are elements of \( { {\mathbb{M}} } \), upper-case hollow font (\( { {\mathbb{U}},{\mathbb{V}},{\mathbb{W}},\ldots } \)) are subsets of \( { {\mathbb{M}} } \). Upper-case Greek letters (\( { \Phi,\Psi,\ldots } \)) are functions on \( { \mathcal{A}^{\mathbb{M}} } \) (e.?g. CA, block maps), and lower-case Greek letters (\( { \phi,\psi,\ldots } \)) are other functions (e.?g. local rules, measures.)
Acronyms in square brackets (e.?g. Topological Dynamics of Cellular Automata) indicate cross‐references to related entries in the Encyclopedia; these are listed at the end of this article.
Bibliography
Akin E (1993) The general topology of dynamical systems, Graduate Studies inMathematics, vol 1. American Mathematical Society, Providence
Allouche JP (1999) Cellular automata, finite automata, and number theory. In:Cellular automata (Saissac, 1996), Math. Appl., vol 460. Kluwer, Dordrecht, pp 321–330
Allouche JP, Skordev G (2003) Remarks on permutive cellular automata. J ComputSyst Sci 67(1):174–182
Allouche JP, von Haeseler F, Peitgen HO, Skordev G (1996) Linear cellularautomata, finite automata and Pascal's triangle. Discret Appl Math 66(1):1–22
Allouche JP, von Haeseler F, Peitgen HO, Petersen A, Skordev G (1997)Automaticity of double sequences generated by one‐dimensional linear cellular automata. Theoret Comput Sci188(1-2):195–209
Barbé A, von Haeseler F, Peitgen HO, Skordev G (1995) Coarse‐graininginvariant patterns of one‐dimensional two-state linear cellular automata. Internat J Bifur Chaos Appl Sci Eng5(6):1611–1631
Barbé A, von Haeseler F, Peitgen HO, Skordev G (2003) Rescaled evolutionsets of linear cellular automata on a cylinder. Internat J Bifur Chaos Appl Sci Eng 13(4):815–842
Belitsky V, Ferrari PA (2005) Invariant measures and convergence properties forcellular automaton 184 and related processes. J Stat Phys 118(3-4):589–623
Blanchard F, Maass A (1997) Dynamical properties of expansive one-sided cellularautomata. Israel J Math 99:149–174
Blank M (2003) Ergodic properties of a simple deterministic traffic flowmodel. J Stat Phys 111(3-4):903–930
Boccara N, Naser J, Roger M (1991) Particle-like structures and theirinteractions in spatiotemporal patterns generated by one‐dimensional deterministic cellular automata. Phys Rev A44(2):866–875
Boyle M, Lind D (1997) Expansive subdynamics. Trans Amer Math Soc349(1):55–102
Boyle M, Fiebig D, Fiebig UR (1997) A dimension group for localhomeomorphisms and endomorphisms of onesided shifts of finite type. J Reine Angew Math 487:27–59
Burton R, Steif JE (1994) Non‐uniqueness of measures of maximal entropyfor subshifts of finite type. Ergodic Theory Dynam Syst 14(2):213–235
Burton R, Steif JE (1995) New results on measures of maximal entropy. Israel JMath 89(1-3):275–300
Cai H, Luo X (1993) Laws of large numbers for a cellular automaton. AnnProbab 21(3):1413–1426
Cattaneo G, Formenti E, Manzini G, Margara L (1997) On ergodic linear cellularautomata over Z m . In: STACS 97 (Lübeck). Lecture Notesin Computer Science, vol 1200. Springer, Berlin, pp 427–438
Cattaneo G, Formenti E, Manzini G, Margara L (2000) Ergodicity, transitivity,and regularity for linear cellular automata over Z m . TheoretComput Sci 233(1-2):147–164
Ceccherini‐Silberstein T, Fiorenzi F, Scarabotti F (2004) The Garden ofEden theorem for cellular automata and for symbolic dynamical systems. In: Random walks and geometry, de Gruyter, Berlin, pp73–108
Ceccherini‐Silberstein TG, Machì A, Scarabotti F (1999) Amenablegroups and cellular automata. Ann Inst Fourier (Grenoble) 49(2):673–685
Courbage M, Kamiński B (2002) On the directional entropy of\( { \mathbb{Z}^2 } \)?actions generatedby cellular automata. Studia Math 153(3):285–295
Courbage M, Kamiński B (2006) Space-time directional Lyapunov exponentsfor cellular automata. J Stat Phys 124(6):1499–1509
Coven EM (1980) Topological entropy of block maps. Proc Amer Math Soc78(4):590–594
Coven EM, Paul ME (1974) Endomorphisms of irreducible subshifts of finitetype. Math Syst Theory 8(2):167–175
Downarowicz T (1997) The royal couple conceals their mutual relationship:a noncoalescent Toeplitz flow. Israel J Math 97:239–251
Durrett R, Steif JE (1991) Some rigorous results for theGreenberg–Hastings model. J Theoret Probab 4(4):669–690
Durrett R, Steif JE (1993) Fixation results for threshold voter systems. AnnProbab 21(1):232–247
Einsiedler M (2004) Invariant subsets and invariant measures for irreducibleactions on zero‐dimensional groups. Bull London Math Soc 36(3):321–331
Einsiedler M (2005) Isomorphism and measure rigidity for algebraic actions onzero‐dimensional groups. Monatsh Math 144(1):39–69
Einsiedler M, Lind D (2004) Algebraic \( { \mathbb{Z} } \) d?actions on entropy rank one. Trans Amer Math Soc 356(5):1799–1831(electronic)
Einsiedler M, Rindler H (2001) Algebraic actions of the discrete Heisenberg group and other non?abelian groups. Aequationes Math 62(1–2):117–135
Einsiedler M, Ward T (2005) Entropy geometry and disjointness forzero‐dimensional algebraic actions. J Reine Angew Math 584:195–214
Einsiedler M, Lind D, Miles R, Ward T (2001) Expansive subdynamics foralgebraic \( { \mathbb{Z} }\) d?actions. Ergodic Theory Dynam Systems21(6):1695–1729
Eloranta K, Nummelin E (1992) The kink of cellular automaton Rule 18 performsa random walk. J Stat Phys 69(5-6):1131–1136
Fagnani F, Margara L (1998) Expansivity, permutivity, and chaos for cellularautomata. Theory Comput Syst 31(6):663–677
Ferrari PA, Maass A, Martínez S, Ney P (2000) Cesàro meandistribution of group automata starting from measures with summable decay. Ergodic Theory Dynam Syst 20(6):1657–1670
Finelli M, Manzini G, Margara L (1998) Lyapunov exponents versus expansivityand sensitivity in cellular automata. J Complexity 14(2):210–233
Fiorenzi F (2000) The Garden of Eden theorem for sofic shifts. Pure Math Appl11(3):471–484
Fiorenzi F (2003) Cellular automata and strongly irreducible shifts of finitetype. Theoret Comput Sci 299(1-3):477–493
Fiorenzi F (2004) Semi-strongly irreducible shifts. Adv Appl Math32(3):421–438
Fisch R (1990) The one‐dimensional cyclic cellular automaton:a system with deterministic dynamics that emulates an interacting particle system with stochastic dynamics. J Theoret Probab3(2):311–338
Fisch R (1992) Clustering in the one‐dimensional three-color cycliccellular automaton. Ann Probab 20(3):1528–1548
Fisch R, Gravner J (1995) One‐dimensional deterministicGreenberg–Hastings models. Complex Systems 9(5):329–348
Furstenberg H (1967) Disjointness in ergodic theory, minimal sets, anda problem in Diophantine approximation. Math Systems Theory 1:1–49
Gilman RH (1987) Classes of linear automata. Ergodic Theory Dynam Syst7(1):105–118
Gottschalk W (1973) Some general dynamical notions. In: Recent advances intopological dynamics (Proc Conf Topological Dynamics, Yale Univ, New Haven, 1972; in honor of Gustav Arnold Hedlund), Lecture Notes in Math, vol 318. Springer, Berlin, pp120–125
Grassberger P (1984) Chaos and diffusion in deterministic cellularautomata. Phys D 10(1-2):52–58, cellular automata (Los Alamos, 1983)
Grassberger P (1984) New mechanism for deterministic diffusion. Phys Rev A28(6):3666–3667
Grigorchuk RI (1984) Degrees of growth of finitely generated groups and thetheory of invariant means. Izv Akad Nauk SSSR Ser Mat 48(5):939–985
Gromov M (1981) Groups of polynomial growth and expanding maps. Inst HautesÉtudes Sci Publ Math 53:53–73
Gromov M (1999) Endomorphisms of symbolic algebraic varieties. J Eur Math Soc(JEMS) 1(2):109–197
von Haeseler F, Peitgen HO, Skordev G (1992) Pascal's triangle, dynamicalsystems and attractors. Ergodic Theory Dynam Systems 12(3):479–486
von Haeseler F, Peitgen HO, Skordev G (1993) Cellular automata, matrixsubstitutions and fractals. Ann Math Artificial Intelligence 8(3-4):345–362, theorem proving and logic programming(1992)
von Haeseler F, Peitgen HO, Skordev G (1995) Global analysis ofself‐similarity features of cellular automata: selected examples. Phys D 86(1-2):64–80, chaos, order and patterns: aspects ofnonlinearity– the “gran finale” (Como, 1993)
von Haeseler F, Peitgen HO, Skordev G (1995) Multifractal decompositions ofrescaled evolution sets of equivariant cellular automata. Random Comput Dynam 3(1-2):93–119
von Haeseler F, Peitgen HO, Skordev G (2001) Self-similar structure of rescaledevolution sets of cellular automata. I. Internat J Bifur Chaos Appl Sci Eng 11(4):913–926
von Haeseler F, Peitgen HO, Skordev G (2001) Self-similar structure of rescaledevolution sets of cellular automata. II. Internat J Bifur Chaos Appl Sci Eng 11(4):927–941
Hedlund GA (1969) Endormorphisms and automorphisms of the shift dynamicalsystem. Math Syst Theory 3:320–375
Hilmy H (1936) Sur les centres d'attraction minimaux des systeémesdynamiques. Compositio Mathematica 3:227–238
Host B (1995) Nombres normaux, entropie, translations. Israel J Math91(1-3):419–428
Host B, Maass A, Martìnez S (2003) Uniform Bernoulli measure in dynamicsof permutative cellular automata with algebraic local rules. Discret Contin Dyn Syst 9(6):1423–1446
Hurd LP, Kari J, Culik K (1992) The topological entropy of cellular automata isuncomputable. Ergodic Theory Dynam Syst 12(2):255–265
Hurley M (1990) Attractors in cellular automata. Ergodic Theory Dynam Syst10(1):131–140
Hurley M (1990) Ergodic aspects of cellular automata. Ergodic Theory Dynam Syst10(4):671–685
Hurley M (1991) Varieties of periodic attractor in cellular automata. TransAmer Math Soc 326(2):701–726
Hurley M (1992) Attractors in restricted cellular automata. Proc Amer Math Soc115(2):563–571
Jen E (1988) Linear cellular automata and recurring sequences in finitefields. Comm Math Phys 119(1):13–28
Johnson A, Rudolph DJ (1995) Convergence under \( { \times_q } \) of \( { \times_p } \) invariant measures on the circle. Adv Math115(1):117–140
Johnson ASA (1992) Measures on the circle invariant under multiplication bya nonlacunary subsemigroup of the integers. Israel J Math 77(1-2):211–240
Kitchens B (2000) Dynamics of \( { mathbf{Z}^d } \) actions on Markov subgroups. In: Topics in symbolic dynamics andapplications (Temuco, 1997), London Math Soc Lecture Note Ser, vol 279. Cambridge Univ Press, Cambridge, pp 89–122
Kitchens B, Schmidt K (1989) Automorphisms of compact groups. Ergodic TheoryDynam Syst 9(4):691–735
Kitchens B, Schmidt K (1992) Markov subgroups of \( { (\mathbf{Z}/2\mathbf{Z})^{\mathbf{Z}^2} } \). In: Symbolicdynamics and its applications (New Haven, 1991), Contemp Math, vol 135. Amer Math Soc, Providence, pp 265–283
Kitchens BP (1987) Expansive dynamics on zero‐dimensional groups. ErgodicTheory Dynam Syst 7(2):249–261
Kleveland R (1997) Mixing properties of one‐dimensional cellularautomata. Proc Amer Math Soc 125(6):1755–1766
Kůrka P (1997) Languages, equicontinuity and attractors in cellularautomata. Ergodic Theory Dynam Systems 17(2):417–433
Kůrka P (2001) Topological dynamics of cellular automata. In: Codes,systems, and graphical models (Minneapolis, 1999), IMA vol Math Appl, vol 123. Springer, New York, pp 447–485
Kůrka P (2003) Cellular automata with vanishing particles. Fund Inform58(3-4):203–221
Kůrka P (2005) On the measure attractor of a cellularautomaton. Discret Contin Dyn Syst (suppl.):524–535
Kůrka P, Maass A (2000) Limit sets of cellular automata associated toprobability measures. J Stat Phys 100(5-6):1031–1047
Kůrka P, Maass A (2002) Stability of subshifts in cellular automata. FundInform 52(1-3):143–155, special issue on cellular automata
Lind D, Marcus B (1995) An introduction to symbolic dynamics and coding.Cambridge University Press, Cambridge
Lind DA (1984) Applications of ergodic theory and sofic systems to cellularautomata. Phys D 10(1-2):36–44, cellular automata (Los Alamos, 1983)
Lind DA (1987) Entropies of automorphisms of a topological Markovshift. Proc Amer Math Soc 99(3):589–595
Lucas E (1878) Sur les congruences des nombres eulériens et lescoefficients différentiels des functions trigonométriques suivant un module premier. Bull Soc Math France6:49–54
Lyons R (1988) On measures simultaneously 2- and 3?invariant. Israel JMath 61(2):219–224
Maass A (1996) Some dynamical properties of one‐dimensional cellularautomata. In: Dynamics of complex interacting systems (Santiago, 1994), Nonlinear Phenom. Complex Systems, vol 2. Kluwer, Dordrecht, pp35–80
Maass A, Martínez S (1998) On Cesàro limit distribution ofa class of permutative cellular automata. J Stat Phys 90(1–2):435–452
Maass A, Martínez S (1999) Time averages for some classes of expansiveone‐dimensional cellular automata. In: Cellular automata and complex systems (Santiago, 1996), Nonlinear Phenom. Complex Systems, vol 3. Kluwer, Dordrecht, pp 37–54
Maass A, Martínez S, Pivato M, Yassawi R (2006) Asymptotic randomizationof subgroup shifts by linear cellular automata. Ergodic Theory Dynam Syst 26(4):1203–1224
Maass A, Martínez S, Pivato M, Yassawi R (2006) Attractiveness of the Haarmeasure for the action of linear cellular automata in abelian topological Markov chains. In: Dynamics and Stochastics: Festschrift in honour of MichaelKeane, vol 48 of, Lecture Notes Monograph Series of the IMS, vol 48. Institute for Mathematical Statistics, Beachwood, pp 100–108
Maass A, Martínez S, Sobottka M (2006) Limit measures for affine cellularautomata on topological Markov subgroups. Nonlinearity 19(9):2137–2147,http://stacks.iop.org/0951-7715/19/2137
Machì A, Mignosi F (1993) Garden of Eden configurations for cellularautomata on Cayley graphs of groups. SIAM J Discret Math 6(1):44–56
Maruoka A, Kimura M (1976) Condition for injectivity of global maps fortessellation automata. Inform Control 32(2):158–162
Mauldin RD, Skordev G (2000) Random linear cellular automata: fractalsassociated with random multiplication of polynomials. Japan J Math (NS) 26(2):381–406
Meester R, Steif JE (2001) Higher‐dimensional subshifts of finite type,factor maps and measures of maximal entropy. Pacific J Math 200(2):497–510
Milnor J (1985) Correction and remarks: “On the concept ofattractor”. Comm Math Phys 102(3):517–519
Milnor J (1985) On the concept of attractor. Comm Math Phys99(2):177–195
Milnor J (1986) Directional entropies of cellular automaton-maps. In:Disordered systems and biological organization (Les Houches, 1985), NATO Adv Sci Inst Ser F Comput Syst Sci, vol 20. Springer, Berlin, pp113–115
Milnor J (1988) On the entropy geometry of cellular automata. Complex Syst2(3):357–385
Miyamoto M (1979) An equilibrium state for a one‐dimensional lifegame. J Math Kyoto Univ 19(3):525–540
Moore C (1997) Quasilinear cellular automata. Phys D103(1–4):100–132, lattice dynamics (Paris, 1995)
Moore C (1998) Predicting nonlinear cellular automata quickly by decomposingthem into linear ones. Phys D 111(1–4):27–41
Moore EF (1963) Machine models of self reproduction. Proc Symp Appl Math14:17–34
Myhill J (1963) The converse of Moore's Garden-of-Eden theorem. Proc AmerMath Soc 14:685–686
Nasu M (1995) Textile systems for endomorphisms and automorphisms of theshift. Mem Amer Math Soc 114(546):viii+215
Nasu M (2002) The dynamics of expansive invertible onesided cellularautomata. Trans Amer Math Soc 354(10):4067–4084 (electronic)
Park KK (1995) Continuity of directional entropy for a class of\( { \mathbf{Z}^2 } \)?actions. J KoreanMath Soc 32(3):573–582
Park KK (1996) Entropy of a skew product with a \( { \mathbf{Z}^2 } \)?action. Pacific J Math172(1):227–241
Park KK (1999) On directional entropy functions. Israel J Math113:243–267
Parry W (1964) Intrinsic Markov chains. Trans Amer Math Soc112:55–66
Pivato M (2003) Multiplicative cellular automata on nilpotent groups:structure, entropy, and asymptotics. J Stat Phys 110(1-2):247–267
Pivato M (2005) Cellular automata versus quasisturmian shifts. Ergodic TheoryDynam Syst 25(5):1583–1632
Pivato M (2005) Invariant measures for bipermutative cellularautomata. Discret Contin Dyn Syst 12(4):723–736
Pivato M (2007) Spectral domain boundaries cellular automata. FundamentaInformaticae 77(special issue), available at: http://arxiv.org/abs/math.DS/0507091
Pivato M (2008) Module shifts and measure rigidity in linear cellular automata. Ergodic Theory Dynam Syst (to appear)
Pivato M, Yassawi R (2002) Limit measures for affine cellularautomata. Ergodic Theory Dynam Syst 22(4):1269–1287
Pivato M, Yassawi R (2004) Limit measures for affine cellularautomata. II. Ergodic Theory Dynam Syst 24(6):1961–1980
Pivato M, Yassawi R (2006) Asymptotic randomization of sofic shifts by linearcellular automata. Ergodic Theory Dynam Syst 26(4):1177–1201
Rudolph DJ (1990) \( { \times 2 } \) and \({ \times 3 } \) invariant measures and entropy. Ergodic Theory Dynam Syst10(2):395–406
Sablik M (2006) Étude de l'action conjointe d'un automate cellulaire etdu décalage: Une approche topologique et ergodique. Ph?D thesis, Université de la Méditerranée, Faculté des science de Luminy,Marseille
Sablik M (2008) Directional dynamics for cellular automata:A sensitivity to initial conditions approach. submitted to Theor Comput Sci 400(1–3):1–18
Sablik M (2008) Measure rigidity for algebraic bipermutative cellularautomata. Ergodic Theory Dynam Syst 27(6):1965–1990
Sato T (1997) Ergodicity of linear cellular automata over \( { \mathbf{Z}_m } \). Inform Process Lett61(3):169–172
Schmidt K (1995) Dynamical systems of algebraic origin, Progress inMathematics, vol 128. Birkhäuser, Basel
Shereshevsky MA (1992) Ergodic properties of certain surjective cellularautomata. Monatsh Math 114(3-4):305–316
Shereshevsky MA (1992) Lyapunov exponents for one‐dimensional cellularautomata. J Nonlinear Sci 2(1):1–8
Shereshevsky MA (1993) Expansiveness, entropy and polynomial growth forgroups acting on subshifts by automorphisms. Indag Math (NS) 4(2):203–210
Shereshevsky MA (1996) On continuous actions commuting with actions ofpositive entropy. Colloq Math 70(2):265–269
Shereshevsky MA (1997) K?property of permutative cellularautomata. Indag Math (NS) 8(3):411–416
Shereshevsky MA, Afra?movich VS (1992/93) Bipermutative cellularautomata are topologically conjugate to the one-sided Bernoulli shift. Random Comput Dynam 1(1):91–98
Shirvani M, Rogers TD (1991) On ergodic one‐dimensional cellularautomata. Comm Math Phys 136(3):599–605
Silberger S (2005) Subshifts of the three dot system. Ergodic Theory DynamSyst 25(5):1673–1687
Smillie J (1988) Properties of the directional entropy function for cellularautomata. In: Dynamical systems (College Park, 1986–87), Lecture Notes in Math, vol 1342. Springer, Berlin, pp689–705
Sobottka M (2005) Representación y aleatorización en sistemasdinámicos de tipo algebraico. Ph?D thesis, Universidad de Chile, Facultad de ciencias físicas y matemáticas,Santiago
Sobottka M (2007) Topological quasi-group shifts. Discret Continuous Dyn Syst17(1):77–93
Sobottka M (to appear 2007) Right‐permutative cellular automata ontopological Markov chains. Discret Continuous Dyn Syst. Available at http://arxiv.org/abs/math/0603326
Steif JE (1994) The threshold voter automaton at a critical point. AnnProbab 22(3):1121–1139
Takahashi S (1990) Cellular automata and multifractals: dimension spectra oflinear cellular automata. Phys D 45(1–3):36–48, cellular automata: theory and experiment (Los Alamos, NM, 1989)
Takahashi S (1992) Self‐similarity of linear cellular automata. J Comput Syst Sci 44(1):114–140
Takahashi S (1993) Cellular automata, fractals and multifractals: space-timepatterns and dimension spectra of linear cellular automata. In: Chaos in Australia (Sydney, 1990), World Sci Publishing, River Edge, pp173–195
Tisseur P (2000) Cellular automata and Lyapunov exponents. Nonlinearity13(5):1547–1560
Walters P (1982) An introduction to ergodic theory, Graduate Texts inMathematics, vol 79. Springer, New York
Weiss B (2000) Sofic groups and dynamical systems. Sankhyā Ser A62(3):350–359
Willson SJ (1975) On the ergodic theory of cellular automata. Math SystTheory 9(2):132–141
Willson SJ (1984) Cellular automata can generate fractals. Discret Appl Math8(1):91–99
Willson SJ (1984) Growth rates and fractional dimensions in cellularautomata. Phys D 10(1-2):69–74, cellular automata (Los Alamos, 1983)
Willson SJ (1986) A use of cellular automata to obtain families offractals. In: Chaotic dynamics and fractals (Atlanta, 1985), Notes Rep Math Sci Eng, vol 2. Academic Press, Orlando, pp123–140
Willson SJ (1987) Computing fractal dimensions for additive cellularautomata. Phys D 24(1-3):190–206
Willson SJ (1987) The equality of fractional dimensions for certain cellularautomata. Phys D 24(1-3):179–189
Wolfram S (1985) Twenty problems in the theory of cellular automata. PhysicaScripta 9:1–35
Wolfram S (1986) Theory and Applications of Cellular Automata. WorldScientific, Singapore
Acknowledgments
I would like to thank François Blanchard, Mike Boyle, Maurice Courbage, Doug Lind, Petr Kůrka, Servet Martínez, Kyewon Koh Park,Mathieu Sablik, Jeffrey Steif, and Marcelo Sobottka, who read draft versions of this article and made many invaluable suggestions, corrections, andcomments. (Any errors which remain are mine.) To Reem.
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2012 Springer-Verlag
About this entry
Cite this entry
Pivato, M. (2012). Ergodic Theory of Cellular Automata. In: Meyers, R. (eds) Computational Complexity. Springer, New York, NY. https://doi.org/10.1007/978-1-4614-1800-9_62
Download citation
DOI: https://doi.org/10.1007/978-1-4614-1800-9_62
Publisher Name: Springer, New York, NY
Print ISBN: 978-1-4614-1799-6
Online ISBN: 978-1-4614-1800-9
eBook Packages: Computer ScienceReference Module Computer Science and Engineering