On the topology of the limit set of non-autonomous IFS Author to whom any correspondence should be addressed.
Yuto Nakajima · 2025
Read On the topology of the limit set of non-autonomous IFS Author to whom any correspondence should be addressed. by Yuto Nakajima (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-2510-23255.txt.
Introduction
Yuto Nakajima and Takayuki Watanabe study the shape of limit sets formed by iterated function systems whose maps change at each step. When rules shift, even at random, the limit shape is poorly known next to the fixed case of Mandelbrot and Hata.
The authors build a framework based on simplices and homology that makes the shape of such limit sets open to strict proof. They then apply the abstract theory to concrete randomly generated systems. Knowing these shapes extends fractal geometry to the more realistic setting where rules vary over time.
Catalog record
| CREATOR | Yuto Nakajima |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 82 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2510-23255
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
On the topology of the limit set of non-autonomous IFS Author to whom any correspondence should be addressed. Yuto Nakajima Abstract Fractals are ubiquitous in nature, and since Mandelbrot’s seminal insight into their structure, there has been growing interest in them. While the topological properties of the limit sets of IFSs have been studied—notably in the pioneering work of Hata—many aspects remain poorly understood, especially in the non-autonomous setting. In this paper, we present a homological framework which captures the structure of the limit set. We apply our novel abstract theory to the concrete analysis of the so-called fractal square, and provide an answer to a variant of Mandelbrot’s percolation problem. This work offers new insights into the topology of fractals. Keywords: non-autonomous IFS, fractal percolation, topology of fractal, fractal squares, Čech (co)homology. MSC2020: 28A80, 55N05, 60K35, 37H12 1 Introduction and the main theorems 1.1 Background Fractals are ubiquitous in nature. For example, the shapes of coastlines, clouds, and forests are typical examples. Mandelbrot pointed out in his seminal work [ 22 ] that these complicated shapes exhibit self-similarity, sometimes only in a statistical sense. The fractals which we study in the present paper are limit sets of non-autonomous iterated function systems, defined below. Definition 1.1 . Let X be a compact metric space. For a map f\colon X\to X , we denote its Lipschitz constant by \mathrm{Lip}(f) . A non-autonomous iterated function system (\Phi^{(j)})_{j=1}^{\infty} on X is a sequence of collections \Phi^{(j)}=\{f_{i}^{(j)}\colon X\to X\}_{i\in I^{(j)}} of maps, where each index set I^{(j)} is finite, and there exists a uniform constant c<1 such that \mathrm{Lip}(f_{i}^{(j)})\leq c for all j\geq 1 and i\in I^{(j)} . For a non-autonomous IFS (\Phi^{(j)})_{j=1}^{\infty} , we endow I^{(j)} with the discrete topology and endow \prod_{j=1}^{\infty}I^{(j)} with the product topology. Define the continuous map \Pi\colon\prod_{j=1}^{\infty}I^{(j)}\to X by \{\Pi(i_{1},i_{2},\dots)\}=\bigcap_{j=1}^{\infty}f^{(1)}_{i_{1}}\circ f^{(2)}_{i_{2}}\circ\dots\circ f^{(j)}_{i_{j}}(X), which is well-defined by the uniform contraction condition. We call \Pi the coding map of (\Phi^{(j)})_{j=1}^{\infty} . Moreover, the image J=\Pi(\prod_{j=1}^{\infty}I^{(j)}) is called the limit set of (\Phi^{(j)})_{j=1}^{\infty} . The concept of a non-autonomous IFS generalizes the classical (autonomous) IFS, where \Phi^{(j)} remains the same for every j\geq 1 . Some of the non-autonomous IFSs considered in this paper are constructed by randomly selecting subsets I^{(j)} from a fixed index set I of an autonomous IFS \{f_{i}\}_{i\in I} . We refer to such non-autonomous systems as random IFSs. As will be reviewed in the subsection 1.3 , much attention has been devoted to the dimension theory of these IFSs. However, the topological properties of fractal sets have not been fully explored. With regard to the topology, Hata [ 13 ] established an equivalent condition for an autonomous limit set to be connected. Bandt and Mesing [ 3 ] studied the topology of IFSs “of finite type”, which offers an interesting viewpoint. For autonomous IFSs of finite type, Luo and Xiong [ 21 ] gave an equivalent condition for the autonomous limit sets to be totally disconnected. For non-autonomous IFSs, Cristea [ 6 ] studied a concrete example of a randomly generated non-autonomous IFS, focusing on when its limit set is connected and when it is totally disconnected. Beyond connectedness, Sumi [ 29 ] went further by generalizing Hata’s result in a simplicial framework and initiated the study of the (co)homology groups of autonomous limit sets. One of the main goals of the present paper is to extend Sumi’s homology theory to the non-autonomous setting. The other main purpose is to study the topological properties of the so-called fractal squares and their generalizations. A typical example is the following. Example 1.2 . Let X=[0,1]^{2} , the unit square, and set I=\{0,1,2\}\times\{0,1,2\} . For each \mathbf{i}=(i_{1},i_{2})\in I , define a contractive map f_{\mathbf{i}}\colon X\to X by f_{\mathbf{i}}(x,y)=\left(\frac{x+i_{1}}{3},\frac{y+i_{2}}{3}\right). For every j\geq 1 , let I^{(j)}\subset I be a non-empty subset. Then \Phi^{(j)}=\{f_{\mathbf{i}}\}_{\mathbf{i}\in I^{(j)}} forms a non-autonomous IFS (\Phi^{(j)})_{j\geq 1} . Call the limit set of Example 1.2 a non-autonomous fractal square. Studying non-autonomous fractal squares is challenging because they do not satisfy a certain condition (see Definition 2.2 ) mentioned in [ 29 ] . Figure 1: Samples of non-autonomous fractal squares. These are randomly constructed as in Theorem 6.18 , with r=1 , 2 , and 3 from left to right. Autonomous fractal squares have been extensively studied in the literature. Their topological properties have been widely investigated, including (total dis)connectedness [ 25 ] , Lipschitz equivalence [ 19 , 26 ] , and Hölder equivalence [ 33 ] . In a notable contribution, Xiao [ 32 ] investigated conditions under which an autonomous fractal square has finitely many connected components. While Xiao’s approach was partly a reworking of Sumi’s theory, he introduced new ideas that enabled a precise count of the number of components. See also the survey [ 20 ] by Luo and Rao. In this paper, we develop a homology theory for non-autonomous IFSs and apply it to the study of non-autonomous fractal squares. We establish conditions under which the limit set is (totally dis)connected. Our result provides an answer to an analog of Mandelbrot’s percolation problem (to be reviewed in the next subsection). We also compute the first Čech homology group \check{H}_{1} of non-autonomous fractal squares, marking the beginning of the study of Fractal Topology. 1.2 Main theorems In Section 2 , we will present a probabilistic result on totally disconnected limit sets generated by random IFSs. Theorem A (Theorem 2.4 ) . Let \{f_{i}\}_{i\in I} be an autonomous IFS which is post-critically finite (see Definition 2.1 ). We choose the index sets I^{(j)}\subset I independently and according to a fixed distribution such that the probability of i\notin I^{(1)} is positive for every i\in I . Then almost surely, the resulting limit set of (\{f_{i}\}_{i\in I^{(j)}})_{j=1}^{\infty} is totally disconnected. In Section 3 , we establish a simplicial framework in which the limit sets of non-autonomous IFSs can be studied. Using a certain kind of self-similarity (Lemma 3.3 ), we construct a nested sequence of simplicial complexes \mathcal{N}_{j,k} (Definition 3.4 ) that capture the overlapping structure of the IFS. The main result here identifies the Čech homology group with the limit of the homology groups of these complexes. Theorem B (Theorem 3.6 and Remark 3.7 ) . Let G be an abelian group and consider homology groups with coefficient G . Then, the group \varprojlim_{k}{H}_{q}(\mathcal{N}_{1,k};G) defined as the inverse limit of homology groups is isomorphic to the Čech homology group \check{H}_{q}(J;G) of the limit set J for every q\geq 0 . Dually, \varinjlim_{k}{H}^{q}(\mathcal{N}_{1,k};G) is isomorphic to the Čech cohomology group \check{H}^{q}(J;G). We call \varprojlim_{k}{H}_{\ast}(\mathcal{N}_{1,k}) and \varinjlim_{k}{H}^{\ast}(\mathcal{N}_{1,k}) the Čech-Sumi homology and cohomology groups for the non-autonomous IFS (\Phi^{(j)})_{j=1}^{\infty} , respectively. In Section 4 , in particular, we establish a correspondence between components of the limit set and components of simplicial complexes. As a consequence, we obtain a Hata-type result and a sufficient condition for total disconnectedness. Theorem C (Corollary 4.8 ) . The following are equivalent. 1. The limit set J is connected. 2. The limit set J is path-connected. 3. For every k>1 , the simplicial complex \mathcal{N}_{1,k} is connected. Theorem D (Corollary 4.10 ) . If \lim_{k\to\infty}c^{k-1}\max\{\#V(\mathcal{K})\colon\mathcal{K}\text{ is a component of }\mathcal{N}_{1,k}\}=0, then the limit set J is totally disconnected. Here, c is the