Welch sets for random generation and representation of reversible one-dimensional cellular automata☆
Introduction
Cellular automata are discrete dynamical systems able to generate complex behavior based on simple local interaction among their components. Reversible cellular automata are characterized to produce an invertible global behavior provoked by a set of local mappings which are not reversible.
In the one-dimensional case, reversible cellular automata have been widely investigated in order to understand their topological, combinatorial and dynamical properties.
Due to their property of keeping the original information of the system, concepts and alternative definitions of reversible cellular automata have been employed in the specification of different cyphering systems [2], [8], [9], [11], [19], [20], [22], and error-correcting codes [13].
Recent works have presented different procedures to characterize reversible behavior in cellular automata. For instance: the generalization of reversibility defined by linear rules over the binary field case under null boundary conditions is explained in [21]. Reversibility of elementary cellular automata rule number 150 is described by circulant matrices in [14]. Reversibility in cellular automata with memory has been studied in [1] and [17].
Thus, there is a continuous investigation to characterize the properties of reversible cellular automata, both for theoretical research and practical application.
The first topological study of reversible cellular automata is developed by Hedlund in [10] establishing important concepts for the one-dimensional case such as the uniform multiplicity of ancestors and Welch sets. Several investigations followed this paper using Welch sets to define new graphs representing the local behavior inducing global reversibility [15], and using their vector representation to give a maximum bound for the minimum information needed to obtain global invertibility [7]. Another topic of interest has been to enumerate all the possible cases of reversible one-dimensional cellular automata [3], [4], [18], the last two references taking reversible automata as groupoids with algebraic properties. In this sense, references [6] and [5] establish the properties of these groupoids and their relation with isotopy and Cayley graphs.
Based on this previous work, the aim of this paper is to present an algorithm to create random Welch sets to define reversible one-dimensional cellular automata. From these Welch sets, we define two bipartite graphs, both for representing the forward and reverse local mappings of the associated reversible cellular automaton. We give examples showing that the former graph provides an alternative representation for the dynamics of reversible automata by block mappings and shifts firstly demonstrated by Kari in [12], and the latter needs less space than the original evolution rule.
The paper is organized as follows: Section 2 explains the basic concepts of reversible one-dimensional cellular automata and Welch sets. Section 3 defines the properties of Welch Sets for reversible automata with neighborhood size 2. Section 4 presents the algorithm generating random Welch sets and arbitrary reversible cellular automata. Section 5 exposes the construction of a bipartite graph based on Welch elements to calculate the evolution of reversible cellular automata and represent their dynamics by block mappings and shifts. Section 6 specifies a shorter version of this graph in order to represent the evolution of reversible automata with a reduced number of elements. The last section gives the final remarks of the study.
Section snippets
Basic concepts of Welch sets
One-dimensional cellular automata are defined by a discrete set of states S, a neighborhood radius an initial configuration for some and an evolution rule . The dynamics is given applying the evolution rule over every cell in c0, thus taking periodic boundary conditions. In general .
For simplicity we shall represent just as φ(c(i)). Thus, for a string v ∈ Sm with we define φ(v) as
Properties of Welch sets
The previous properties give a particular structure for Mφ and . Property 1 guarantees that there are |S| entries equal to s in both matrices for every s ∈ S. Property 2 establishes that if then for si ∈ S. Finally, Property 3 specifies that for every state s ∈ S there is a unique state a ∈ S such that . Thus, in Mφ and every state has a rectangular arrangement, and both matrices define a rectangular groupoid [5].
For every reversible
Algorithm to generate random Welch sets
For U⊆S, we can represent the elements of U by a binary vector gU: S → 0, 1 of size 1 × |S| such that iff a ∈ U; otherwise, . With this, we define two mappings and .
The next algorithm computes valid Welch sets for every state s ∈ S to fill up the matrix Mφ representing the evolution rule of a reversible cellular automaton. The algorithm applies the properties of Welch sets as vector and matrix operations in order to facilitate their computational implementation.
First representation: bipartite graph with Welch elements
For s ∈ S and its left Welch set Ls, a left Welch element is a pair (a, s) such that a ∈ Ls. Analogously, a right Welch element is a pair (s, a) such that a ∈ Rs.
Given the matrices GL, GR and Mφ, we can define a bipartite graph where NL is the set of left Welch elements and NR is the set of right Welch elements for every state in S.
Every left Welch element (a, b) ∈ NL holds that a ∈ Lb for a, b ∈ S. Analogously, every right Welch element (d, e) ∈ NR holds that e ∈ Rd for d, e ∈ S. There
Second representation: Bipartite graph with welch sets
With the matrices GL, GR and Mφ, another bipartite graph is defined where is the family of left Welch sets and is the family of right Welch sets. For every s ∈ S, each set and every . There is a directed edge (Ls, Rs) labeled by s. The directed edges from into are implicitly defined by Property 3. That is, for each and every .
Fig. 9 presents de digraphs Q associated with the reversible automata previously used in Fig. 4. Bold lines describe the
Final remarks
This paper has presented an algorithm based on the properties of Welch indices in order to calculate a random reversible cellular automaton. This algorithm is able to calculate reversible automata with tens of states.
The properties of Welch indices are employed as well to specify two types of bipartite graphs, which represent the dynamics of reversible automata.
The first graph takes as nodes the Welch elements, and the edges among them are utilized to represent the evolution in both directions
Acknowledgment
This work has been supported by CONACYT project No. CB-2014-237323 and by IPN Collaboration Network “ Grupo de Sistemas Complejos del IPN ”.
References (22)
- et al.
An authenticated image encryption scheme based on chaotic maps and memory cellular automata
Opt. Lasers Eng.
(2013) Efficient exhaustive listings of reversible one dimensional cellular automata
Theor. Comput. Sci.
(2004)Rectangular groupoids and related structures
Discrete Math.
(2013)Design of fast one-pass authenticated and randomized encryption schema using reversible cellular automata
Commun. Nonlin. Sci. Numer. Simul.
(2014)Fast encryption of rgb color digital images using a tweakable cellular automaton based schema
Opt. Laser Technol.
(2014)- et al.
Randomized local search, evolutionary algorithms, and the minimum spanning tree problem
Theor. Comput. Sci.
(2007) - et al.
Invertible behavior in elementary cellular automata with memory
Inf. Sci.
(2012) - et al.
Procedures for calculating reversible one-dimensional cellular automata
Physica D
(2005) - et al.
A novel image encryption algorithm using chaos and reversible cellular automata
Commun. Nonlin. Sci. Numer. Simul.
(2013) - et al.
Reversibility of general 1d linear cellular automata over the binary field z2 under null boundary conditions
Inf. Sci.
(2015)
Reversible cellular automata with memory of delay type
Complexity
Cited by (10)
A closed formula for the inverse of a reversible cellular automaton with (2R+1)-cyclic rule
2019, Applied Mathematics and ComputationCitation Excerpt :In this sense, it is shown that such algorithm exists for the case of uni-dimensional cellular automata [1], whereas this question is undecidable for two-dimensional cellular automata [12]. Since these earlier studies there have been a number of works dealing with the study of reversibility properties of cellular automata; in fact it remains a very topical issue (see, for example, [2,4,9,14,26,27]). For a more detailed explanation of reversible cellular automata we refer the reader to [13,20,21] and references therein.
Brief notes and history of computing in Mexico during 50 years
2020, International Journal of Parallel, Emergent and Distributed Systems
- ☆
Dedicated to the memory of Harold V. McIntosh (1929–2015), a leader in the development of cellular automata theory, whose teaching and knowledge were essential to this work.