Open access peer-reviewed chapter

# Partial Entropy and Bundle-Like Entropy for Topological Dynamical Systems

Written By

Kesong Yan and Fanping Zeng

Submitted: May 14th, 2019 Reviewed: August 5th, 2019 Published: September 23rd, 2019

DOI: 10.5772/intechopen.89021

From the Edited Volume

## Dynamical Systems Theory

Edited by Jan Awrejcewicz and Dariusz Grzelczyk

Chapter metrics overview

View Full Metrics

## Abstract

Entropy is an important notion for understanding the complexity of dynamical systems. Several important entropy-like invariants based on the preimage structure for noninvertible maps have been defined and studied by some authors. In this chapter, following the idea of Hurley, we first further study the relationship among the topological entropy, pseudo-orbit, and preimage entropies for topological dynamical systems from the view of localization. Secondly, two entropy-like invariants, which are called the partial entropy and bundle-like entropy, for nonautonomous discrete dynamical systems are introduced. A relationship between the topological entropy and such two entropies is established.

### Keywords

• topological entropy
• point entropy
• pseudo-orbit
• partial entropy
• bundle-like entropy
• 2000 Mathematics Subject Classification: Primary: 37B40
• 37A35
• 37B10
• 37A05

## 1. Introduction

By a topological dynamical system, we mean a pair X T , where X is a compact metric space with a metric d and T is a continuous surjective map from X to itself [1]. An important notion for understanding the complexity of dynamical systems is topological entropy, which was first introduced by Adler et al. [2] in 1965, and later Dinaburg [3] and Bowen [4] gave two equivalent definitions on a metric space by using separated sets and spanning sets. Roughly speaking, topological entropy measures the maximal exponential growth rate of orbits for an arbitrary topological dynamical system.

When a considered mapping T is invertible, it is well-known that T and the inverse mapping T 1 have the same topological entropy. However, when the map T is not invertible, the “inverse” is set-valued, yielding the iterated preimage set T n x = z X : T n z = x of a point x X which is in general a set rather than a point, so different ways of “extending the procedure into the past” lead to several new entropy-like invariants for non-invertible maps.

In 1991, Langevin and Walczak [5] regard the “inverse” as a relation and formulate a notion of entropy for this relation (analogous to the entropy of a foliation [6]), based on distinguishing points by means of the structure of their “preimage trees,” which is called preimage relation entropy. The interested reader can see [7] or [8] for more details on this invariant. Later, several important entropy-like invariants based on the preimage structure for non-invertible maps, such as pointwise preimage entropies, preimage branch entropy [1, 8, 9, 10], partial preimage entropy, conditional preimage entropy [11], etc., have been introduced, and their relationships with topological entropy have been established. To learn more about the results related to the preimage entropy for noninvertible maps, one can see [12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23].

The local entropy theory for topological dynamical systems started in the early 1990s with the work of Blanchard (see [24, 25]). Nowadays this theory has become a very interesting topic in the field of dynamical systems and has also proven to be fundamental to many other related fields. For example, Blanchard defined the notion of entropy pairs and used it to obtain a disjointness theorem [26]. The notion of entropy pairs can also be used to show the existence of the maximal zero-entropy factor, called the topological Pinsker factor, for any topological dynamical system [25]. In order to gain a better understanding of the topological version of a K-system, the theory of entropy tuples [27, 28, 29] was developed. To learn more about the theory related to the local entropy, we refer the interested reader to see the survey paper [30] and references therein.

We remark that in reality, it is difficult to find a real orbit in the system, but a pseudo-orbit can be used to approximate the real orbit, and so there have been a lot of applications in many fields. Since the works of Bowen [31] and Conley [32], pseudo-orbits have proved to be a powerful tool in dynamical systems. For instance, Hammel et al. [33, 34] have investigated the role of pseudo-orbits in computer simulations of certain dynamical systems; Barge and Swanson [35] made use of pseudo-orbits to study rotation sets of circle and annulus maps. Also, a remarkable result by Misiurewicz [36] showed that the topological entropy can be computed by measuring the exponential growth rate of the numbers of pseudo-orbits (related results can see [37]). In [1], Hurley showed that the point entropy with pseudo-orbits that is defined by replacing inverse orbit segments by inverse pseudo-orbit segments in the definition of pointwise preimage entropy is in fact equal to the topological entropy.

In this chapter, following Hurley [1] we further study the preimage entropy for topological dynamical system from the view of localization. In Section 2, we consider the calculation of topological entropy for open covers from pseudo-orbits (Theorem 2.3). In Section 3, we investigate the relationship among the topological entropy for open covers and several preimage entropy invariants, which is viewed as the local version of the Hurley inequality (Theorem 3.1). In Section 4, we show that the topological entropy for open covers can be computed by measuring the exponential growth rate of the number of pseudo-orbits that end at a particular point (Theorems 4.2 and 4.3).

A nonautonomous discrete dynamical system is a natural generalization of a classical dynamical system; its dynamics is determined by a sequence of continuous self-maps f n : X n X n + 1 , which defined on a sequence on compact metric spaces (Xn , dn ). The topological entropy of nonautonomous discrete dynamical systems was introduced by Kolyada and Snoha [38]. In Section 5, following the idea of [1, 39], we introduce two entropy-like invariants, which are called the partial entropy and bundle-like entropy, for nonautonomous discrete dynamical systems, and study the relationship among them and the topological entropy (Theorems 5.2, 5.3, and 5.5).

## 2. Topological entropy and pseudo-orbits

### 2.1 Topological entropy via open covers

Topological entropy was defined originally by Adler et al. [2] for continuous maps on compact topological spaces. Let X T be a topological dynamical system. A finite open cover of X is a finite family of open sets whose union is X . Denoted by C X o is the set of finite open covers of X . Given two open covers U , V C X o , U is said to be finer than V ( U V ) if each element of U is contained in some element of V . Let U V = U V : U U V V . It is clear that U V U and U V V .

Let U C X o . For two nonnegative integers M N , denoted by U M N = n = M N T n U , where T n U = T n U : U U for all positive integers n . For any K X , define N U K as the minimal cardinality of any subcovers of U that covers K . We write N U X simply by N U . The topological entropy of U with respect to T is defined by

h top T U = lim n 1 n log N U 0 n 1 = inf n 1 1 n log N U 0 n 1 . E1

The topological entropy of T is

h top T = sup U C X o h top T U . E2

### 2.2 Separated sets, spanning sets, and topological entropy

In this subsection, we recall two equivalent definitions, which are given by Dinaburg [3] and Bowen [4]. Let X T be a topological dynamical system. Given a nonempty subset K of X, for any ϵ > 0 and n , a subset E of K is called an n ϵ -separated set of K if any x y E implies d n x y ϵ , where

d n x y max 0 i n 1 d T i x T i y .

Denote the maximal cardinality of any n ϵ -separated subset of K by s n ϵ K . A subset F of K is called an n ϵ -spanning set of K , if for any x K , there exists y F with d n x y < ϵ . Denote the minimal cardinality of any n ϵ -spanning set for K by r n ϵ K .

The following lemma is well-known, and its proof is not difficult, so we omit its detail proof.

Lemma 2.1. Let X T be a topological dynamical system. For any subset K of X and any integer n 1 , we have the following properties:

1. r n ϵ K s n ϵ K r n ϵ / 2 K for all ϵ > 0 .

2. N U 0 n 1 K r n δ K for any n and any U C X o with the Lebesgue number 2 δ .

3. s n ϵ K N U 0 n 1 K for any U C X o with diam U < ϵ .

By Lemma 2.1, we obtain directly the following result.

Theorem 2.2. (see [3, 4, 40]). Let X T be a topological dynamical system. Then

h top T = lim ϵ 0 lim sup n 1 n log s n ϵ X = lim ϵ 0 lim sup n 1 n log r n ϵ X .

### 2.3 Topological entropy via pseudo-orbits

Let X d be a compact metric space. Denote X n as the n -fold Cartesian product of X ( n 1 ) . Fixing a positive number ϵ , a subset E X n is said to be n ϵ -separated if for any two distinct points x ˜ = x 0 x 1 x n 1 , y ˜ = y 0 y 1 y n 1 E , there is a 0 i n 1 such that d x i y i > ϵ . By the compactness of X , any n ϵ -separated set is finite. If Z X n is a nonempty subset, then we denote the maximal cardinality of any n ϵ -separated subset of Z by s n ϵ Z .

Let Z X n be a nonempty subset. A subset F Z is called n ϵ -panning for Z if for each z ˜ = z 0 z 1 z n 1 Z , there is a y ˜ = y 0 y 1 y n 1 F with d z i y i < ϵ for every 0 i n 1 . We denote the minimal cardinality of any n ϵ -spanning subset of Z by r n ϵ Z .

For each positive integer n 1 , we let O n denote the set of all orbit segments of length n , that is,

O n = x Tx T n 1 x X n : x X .

Note that a point w ˜ = x Tx T n 1 x O n is uniquely determined by its initial point x X . Thus, we have

h top T = lim ϵ 0 lim α 0 lim sup n 1 n log s n ϵ O n = lim ϵ 0 lim α 0 lim sup n 1 n log r n ϵ O n .

Topological entropy has been characterized by Misiurewicz [36] and Barge and Swanson [37] in terms of growth rates of pseudo-orbits. Let X T be a topological dynamical system. For α > 0 , an α -pseudo-orbit for T of length n is a point x ˜ = x 0 x 1 x n 1 X n with the property that d T x j 1 x j < α for all 1 j n 1 . Let Ψ n α X n denote all α -pseudo-orbits of length n . It was shown in [36, 37] that

h top T = lim ϵ 0 lim α 0 lim sup n 1 n log s n ϵ Ψ n α = lim ϵ 0 lim α 0 lim sup n 1 n log r n ϵ Ψ n α .

In the following, we will show that the topological entropy for an open cover can be characterized by pseudo-orbits. Before proceeding, let us first introduce a definition of pseudo-orbit entropy via open covers. Let X T be a topological dynamical system. For each integer n 1 and U C X o , we define an open cover U n of the product space X n by

U n = U 0 × U 1 × × U n 1 : U j U for each j = 0 1 n 1 ,

where

U 0 × U 1 × × U n 1 = u 0 u 1 u n 1 : u j U j for each j = 0 1 n 1 .

Given α > 0 , it is not hard to see that a n = N U n Ψ n α is a nonnegative sub-additive sequence, that is, a n + m a n + a m for all positive integers n and m . The α -pseudo-orbit entropy of U is then defined by

h Ψ T U α = lim n 1 n log N U n Ψ n α = inf n 1 1 n log N U n Ψ n α , E3

and the pseudo-orbit entropy of U is defined by

h Ψ T U = lim α 0 h Ψ T U α . E4

Theorem 2.3. Let X T be a topological dynamical system. If U C X o , then we have

h top T U = h Ψ T U . E5

Proof. To prove (5), it suffices to note that h Ψ T U α 1 h Ψ T U α 2 whenever α 1 < α 2 and inf 0 < α 1 N U n Ψ n α = N U n O n = N U 0 n 1 . Thus, we have

h Ψ T U = lim α 0 h Ψ T U α = inf 0 < α 0 inf n 1 1 n log N U n Ψ n α = inf n 1 inf 0 < α 1 1 n log N U n Ψ n α = inf n 1 1 n log N U 0 n 1 = h top T U .

This completes the proof of the theorem. □

Remark 2.4. Combining (2) and (5), we have

h top T = sup U C X o h Ψ T U .

On the other hand, let us define h Ψ T = sup U C X o h Ψ T U , which is called the pseudo-orbit entropy of T . Using the same techniques of topological entropy (see Lemma 2.1), we can easily show that

h Ψ T = lim ϵ 0 lim α 0 lim sup n 1 n log s n ϵ Ψ n α = lim ϵ 0 lim α 0 limsup n 1 n log r n ϵ Ψ n α .

So, it is in fact to give a simpler proof of Theorem 1 of [37] by Theorem 2.3.

## 3. Pointwise preimage entropies for open covers and local Hurley’s inequality

When T is not invertible, one can ask about growth rates of inverse images f n x . In this section we describe two ways of doing this, which were introduced by Hurley in [1].

### 3.1 Preimage branch entropy

Let X T be a topological dynamical system. Given x X let T n x denote the tree of inverse images of x up to order n , which is defined by

T n x = z 0 z 1 z n : z n = x and z j = T z j 1 for all 1 j n .

Each z 0 z 1 z n T n x is called a branch of T n x , and its length is n . Note that every branch of T n x ends with x . Let T n = x X T n x ; we define a metric on T n as follows: suppose that z ˜ = z 0 z 1 z n and w ˜ = w 0 w 1 w n are two branches of the length n , the branch distance between them is defined as

d B , n z ˜ w ˜ = max 0 j n d z j w j .

Let O n = T n x : x X . Given two trees T n x and T n y in O n , the branch Hausdorff distance between them, d bH T n x T n y is the usual Hausdorff metric based upon d B , n ; that is,

d bH T n x T n y = max max z ˜ T n x min w ˜ T n y d B , n z ˜ w ˜ max w ˜ T n y min z ˜ T n x d B , n z ˜ w ˜ .

Note that d bH T n x T n y < ϵ if and only if each branch of either tree is d B , n within ϵ of at least one branch of the other tree. Two trees T n x and T n y in O n are said to be d bH - n ϵ -separated if d bH T n x T n y < ϵ , that is, there is a branch z ˜ of one of the trees with the property that d B , n z ˜ w ˜ > ϵ for all branches w ˜ of the other tree. Let t n ϵ denote the maximum cardinality of any d bH - n ϵ -separated sets of O n . Define the entropy by

h b T = lim ϵ 0 lim sup n 1 n log t n ϵ ,

which is called the preimage branch entropy of T.

### 3.2 Pointwise preimage entropies

Let us recall two non-invertible invariants defined by Hurley [1] in 1995. Hurley’s invariants are about the maximum rate of dispersal of the preimage sets of individual points, which are called pointwise preimage entropies in [8]. The difference between these two invariants is when the maximization takes place:

h p T = sup x X lim ϵ 0 lim sup n 1 n log s n ϵ T n x = sup x X lim ϵ 0 lim sup n 1 n log r n ϵ T n x , h m T = lim ϵ 0 lim sup n 1 n log sup x X s n ϵ T n x = lim ϵ 0 lim sup n 1 n log sup x X r n ϵ T n x .

It is clear that h p T h m T , and in [18] the authors constructed an example for which h p T < h m T . In addition, Hurley established the following relationships among preimage branch entropy, pointwise preimage entropy, and topological entropy (see [1], Theorem 3.1):

h m T h top T h m T + h b T .

We call it the Hurley inequality.

### 3.3 Local Hurley’s inequality

In this subsection, we mainly study the relationship among the topological entropy for open covers and several preimage entropy invariants, which is viewed as the local version of the Hurley inequality. To do it, we first introduced a definition of preimage entropy via open covers.

Let X T be a topological dynamical system. Given U C X o , define two pointwise preimage entropies of U with respect to T by

h p T U = sup x X lim sup n 1 n log N U 0 n 1 T n x

and

h m T U = lim sup n 1 n log sup x X N ( U 0 n 1 T n x ) .

Theorem 3.1. (Local Hurley’s inequality). Let X T be a topological dynamical system. If U C X o , then we have

h p T U h m T U h top T U h m T U + h b T .

Proof. It is obvious that N U 0 n 1 T n x N U 0 n 1 for every x X and every integer n 1 . So that h p T U h m T U h top T U . Now we show the last inequality h top T U h m T U + h b T .

Let ϵ > 0 be a Lebesgue number of U . Fix n 1 , and let Y denote a d bH - n ϵ / 3 -separated set of O n with cardinality t n ϵ / 3 . Let Z denote the set of all root points of trees in Y , where the root point of the tree T n x is x . For each z Z , let V z U be a subcover of U 0 n 1 with cardinality N U 0 n 1 T n z that covers T n z , and let

V = z Z V z U .

We claim that V is an open cover of X .

In fact, let x X be given and let w = f n x . Since Y is a d bH - n ϵ / 3 -separated set of O n with maximal cardinality, there is a tree T n y Y such that d bH T n w T n y < ϵ / 3 . Now we consider the branch w ˜ of T n w begins with x , i.e., w ˜ = x f x f n 1 x f n x = w T n w . Then there exists a branch y ˜ = y 0 y 1 y n = y T n y such that d B , n w ˜ y ˜ < ϵ / 3 . This means that d T j y 0 T j x < ϵ / 3 for each 0 j n . Thus, there exists V V y U such that x V . This yields the claim that V is an open cover of X . So that N U 0 n 1 V , where V denotes the cardinality of V . Using the claim, we have

N U 0 n 1 V z Z V z U = z Z N U 0 n 1 T n z Z sup x X N ( U 0 n 1 T n x ) = Y sup x X N ( U 0 n 1 T n x ) = t n ϵ / 3 sup x X N ( U 0 n 1 T n x ) .

So that,

h top T U = lim n 1 n log N U 0 n 1 lim sup n 1 n log t n ϵ / 3 + log sup x X N ( U 0 n 1 T n x ) lim sup n 1 n log t n ϵ / 3 + lim sup n 1 n log sup x X N ( U 0 n 1 T n x ) = lim sup n 1 n log t n ϵ / 3 + h m T U h b T + h m T U .

This completes the proof of the theorem.□

We remark that Theorem 3.1 generalizes the classical Hurley’s inequality given in [26, Theorem 3.1]. A direct consequence of Theorem 3.1 is.

Corollary 3.2. (Hurley’s inequality). Let X T be a topological dynamical system. Then we have

h p T h m T h top T h m T + h b T . E6

Proof. It follows directly from Lemma 2.1 that

h p T = sup U C X o h p T U and h m T = sup U C X o h p T U . E7

Thus, combining (2), (7), and Theorem 3.1 gives (6).□

## 4. Point entropy for open covers with pseudo-orbits

In [1], Hurley considered pseudo-orbits for inverse images and showed that the topological entropy can be characterized in terms of growth rates of pseudo-orbits that end at a particular point. Let X T be a topological dynamical system. Recall that if α > 0 , then an α -pseudo-orbit x 0 x 1 x n 1 X n is an approximate orbits segment for T in the sense that d T x j x j + 1 < α for all 0 j n 1 .

For each x X , let Ψ n α x X n denote the set of all α -pseudo-orbits of length n that end at x , i.e., an element of Ψ n α x is an α -pseudo-orbit y 0 y 1 y n 1 with y n 1 = x . It was shown in [1], (Propositions 4.2 and 4.3) that

h top = lim ϵ 0 lim α 0 lim sup n 1 n log max x X s ( n ϵ Ψ n α x ) = sup x X lim ϵ 0 lim α 0 lim sup n 1 n log s n ϵ Ψ n α x . E8

In either formula s n ϵ Ψ n α x can be replaced by r n ϵ Ψ n α x .

In the following, we will show that the topological entropy for an open cover can be characterized by pseudo-orbits for inverse images. Before proceeding, let us consider the following definitions, which use the notation introduced in Section 2.3.

Let X T be a topological dynamical system. For each integer n 1 , U C X o , and α > 0 , we define

N max n U α = max x X N U n Ψ n α x . E9

Clearly,

N U n Ψ n α x N max n U α N U n Ψ n α E10

for every x X . In addition, by the compactness of X , there is some point y X such that

N U n Ψ n α y = N max n U α .

Lemma 4.1. Let X T be a topological dynamical system and U C X o . Suppose that ε > 0 is a Lebesgue number of U and 0 < α < ε / 4 . Then there is a constant K = K α such that for every n 1 ,

N U n Ψ n α K N max n U α . E11

Proof. Let x 1 x 2 x K be a finite α -dense subset of X , i.e., i = 1 n B x i α = X , where B x i α = z X : d x i z < α . For each 1 i K , let V i be a subcover of U n that covers Ψ n α x i with cardinality N U n Ψ n α x i . Define V = i = 1 K V i . Clearly, V i = 1 K V i K N max n U α . So, to complete the proof of the lemma, it suffices to show V is a subcover of U n that covers Ψ n α .

In fact, let y ˜ = y 0 y 1 y n 1 be an α -pseudo-orbit. Since x 1 x 2 x K is an α -dense subset of X , there is some x i satisfying d T y n 2 x i < α . This implies z ˜ = z 0 z 1 z n 2 z n 1 = y 0 y 1 y n 2 x i is an α -pseudo-orbit ending at x i . Since V i is a subcover of U n that covers Ψ n α x i , there is some V V i such that z ˜ V . Since z j = y j for all 0 j n 2 and ϵ is the Lebesgue number of U , in order to show that y ˜ V , we need only to show that d y n 1 x i < ϵ / 2 ; this is obviously, as d y n 1 x i d y n 1 T y n 2 + d T y n 2 x i < 2 α < ϵ / 2 .□

Theorem 4.2. Let X T be a topological dynamical system. If U C X o , then we have

h top T U = lim α 0 lim sup n 1 n log sup x X N ( U n Ψ n α x ) . E12

Proof. Combining (10) and (11), we have

N max n U N U n Ψ n α K N max n U

for each fixed 0 < α < ϵ / 4 and all n 1 , where ϵ is a Lebesgue number of U and K = K α in Lemma 4.1 is independent of n . This implies that

lim sup n 1 n log N max n U α = lim sup n 1 n log N U n Ψ α E13

for all positive number 0 < α < ϵ / 2 . Thus, we have

h top T U = lim α 0 lim n 1 n log N U n Ψ n α by Theorem 2.3 = lim α 0 lim sup n 1 n log N U n Ψ n α = lim α 0 lim sup n 1 n log N max n U α by 4.6 = lim α 0 lim sup n 1 n log sup x X N ( U n Ψ n α x ) by 4.1

This completes the proof.□

Theorem 4.3. Let X T be a topological dynamical system. If U C X o , then we have

h top T U = sup x X lim α 0 lim sup n 1 n log N U n Ψ n α x . E14

Proof. It follows directly from (10) and (12) that

h top T U = lim α 0 lim sup n 1 n log sup x X N ( U n Ψ n α x ) sup x X lim α 0 lim sup n 1 n log N U n Ψ n α x .

Now we start to prove the converse inequality.

Note that for the given n 1 and α > 0 , there is a point y n U α X such that

N ( U n Ψ n α y n U α = max x X N U n Ψ n α x .

Taking a sequence of integers n i = n i α such that

lim sup n 1 n log N U n Ψ n α y n U α = lim i 1 n i log N U n i Ψ n i α y n i U α .

By restricting to a subsequence, we can assume without loss of generality that the sequence y i α = y n i U α converses to a limit q α .

Let ϵ be a Lebesgue number of U . If 0 < β < ϵ / 4 and d y i α q α < β , then V is a subcover of U n that covers Ψ n α y i α whenever V is a subcover of U n that covers Ψ n α + β q α . This implies that

N U n Ψ n α + β q α N U n Ψ n α y i α E15

whenever d y i α q α < β .

Now we choose a sequence α j 0 such that q α j converges to some point q X . Similar to the proof as above we have

N ( U n Ψ n α j + 2 β q N U n Ψ n α j q α j E16

whenever d q α j q < β . Combining inequalities (15) and (16), one has

N U n Ψ n α j + 2 β q N U n Ψ n α j y i α j E17

whenever d y i α j q α j < β and d q α j q < β . If j is a fixed integer with d q α j q < β , then (17) holds for all sufficiently large integers i . Thus,

limsup n 1 n log N U n Ψ n α j + 2 β q limsup i 1 n i log N U n i Ψ n i α j + 2 β q lim i 1 n i log N U n i Ψ n i α j y i α j = limsup n 1 n log max x X N ( U n Ψ n α j x ) ) . E18

Now let j and use the fact that both sides (18) are nonincreasing as α decreases to conclude that

lim sup n 1 n log N U n Ψ n 3 β q lim j 0 lim sup n 1 n log N U n Ψ n α j + 2 β q lim j lim sup n 1 n log max x X N ( U n Ψ n α j x ) ) = inf α > 0 lim sup n 1 n log max x X N ( U n Ψ n α x ) = lim α 0 lim sup n 1 n log max x X N ( U n Ψ n α x ) ) . E19

Therefore, combining (12) and (19), we have

h top T U = lim α 0 limsup n 1 n log max x X N ( U n Ψ n α x ) ) inf β > 0 limsup n 1 n log N U n Ψ n 3 β q = inf α > 0 limsup n 1 n log N U n Ψ n α q = lim α 0 limsup n 1 n log N U n Ψ n α q sup x X lim α 0 limsup n 1 n log N U n Ψ n α x . E20

This completes the proof.□

## 5. Partial entropy and bundle-like entropy for nonautonomous discrete dynamical systems

In [38, 41], topological entropy for certain nonautonomous discrete dynamical system was defined and studied. In this section, we study the topological entropy for nonautonomous discrete dynamical systems by introducing two entropy-like invariants called the partial entropy and bundle-like entropy as being motivated by the idea of [1, 39].

### 5.1 Topological entropy for nonautonomous discrete dynamical systems

Let X be a collection of countable infinitely many compact metric space X i d i and F be a collection of countable infinite many continuous maps f i : X i X i + 1 , i = 1 , 2 , . Then the pair X F is called a nonautonomous discrete dynamical system.

For any integer n 1 , we define a metric d ˜ n on i = 1 n X i as follows: for any two points x ˜ n = x 1 x 2 x n , y ˜ n = y 1 y 2 y n i = 1 n X i ,

d ˜ n x ˜ n y ˜ n = max 1 i n d i x i y i .

Fixing an integer n 1 and a positive number ϵ . A subset Z of i = 1 n X i is called d ˜ n - n ϵ -separated if for any two distinct points x ˜ n , y ˜ n Z we have d ˜ n x ˜ n y ˜ n > ϵ . Denote the maximal cardinality of any d ˜ n -separated subset of Z by s n ϵ Z . A subset W Z is called d ˜ n - n ϵ -spanning for Z if for each z ˜ n Z , there is a w ˜ n W such that d ˜ n z ˜ n w ˜ n < ϵ . Denote the minimal cardinality of any d ˜ n -spanning subset of Z by r n ϵ Z .

The following result is trivial, so we omit its detail proof.

Lemma 5.1. Suppose that n is a positive integer and Z is a nonempty subset of i = 1 n X i . Then for each ϵ > 0 , we have

r n ϵ Z s n ϵ Z r n ϵ / 2 Z .

For each n 1 let Z n be a nonempty subset of i = 1 n X i . Then it follows immediately from Lemma 5.1 that

lim ϵ 0 lim sup n 1 n log r n ϵ Z n = lim ϵ 0 lim sup n 1 n log s n ϵ Z n . E21

Given a nonautonomous discrete dynamical system X F , denoted by O n , F or O n for short the set of all orbit segments of length n for each n 1 , i.e.,

O n = O n , F x 1 x 2 x n : x 1 X 1 and x i + 1 = f i x i i = 1 2 n 1 .

Then the common limit in (21) by taking Z n = O n is defined to be the topological entropy of X F , written h top X F or h top F for short if there is no confusion.

### 5.2 Partial entropy and bundle-like entropy

Let X F be a nonautonomous discrete dynamical system. A collection P = P i : i 1 is said to be a cover of X if each P i covers X i , respectively. We now define two entropies, partial entropy and bundle-like entropy, for X F relative to P .

For any integer n 1 and D P n , let W n D i = 1 n X i denote the set of all orbit segments of length that end at some point x n D , i.e.,

W n D = x 1 x 2 x n O n : x n D .

Put s max , P n n ϵ = sup D P n s n ϵ W n D . Define the entropy by

h p , P X F = lim ϵ 0 lim sup n 1 n log s max , P n n ϵ ,

which is called the partial entropy of X F relative to P and written shortly by h p , P F if there is no confusion.

Let O n , P n = W n D : D P n . For any two elements, W n D and W n E of O n , P n , denoted by d H W n D W n E , the usual Hausdorff metric between them is based upon metric d ˜ n of i = 1 n X i defined as before and by s n ϵ O n , P n the maximum cardinality of any d H - n ϵ -separated subset of O n , P n . Define the entropy by

h b , P X F = lim ϵ 0 lim sup n 1 n log s n ϵ O n , P n ,

which is called the bundle-like entropy of X F relative to P and written shortly by h b , P F if there is no confusion.

Also, we have the spanning set versions of definitions of h p , P F and h b , P F , respectively.

### 5.3 Some relationships between h top F and h p , P F

Theorem 5.2. Let X F be a nonautonomous discrete dynamical system, and P = P i : i 1 be a cover of X . Then we have

h p , P F h top F h b , P F + h p , P F .

Proof. Note that s max , P n n ϵ s n ϵ O n for any cover P of X and any ϵ > 0 . Then the former inequality is obtained. Now we show the later one. If h b , P F = , then there is nothing to prove. Now assuming h b , P F < .

Fixing a sufficiently small ϵ > 0 and an integer n 1 , let Y be a d H - n ϵ -separated subset of O n , P n with cardinality s n ϵ O n , P n . For each W n D Y , let M D be a d ˜ n - n ϵ -separated subset of W n D with cardinality s n ϵ W n D . Put M = W n D Y M D . We claim that M is a d ˜ n - n 3 ϵ -spanning subset of O n .

In fact, for any x = x 1 x 2 x n O n , since Y is a d H - n ϵ -separated subset of O n , P n with maximum cardinality and P n covers X n , there is an E P n with x n E and a W n D Y such that d H W n D W n E ϵ . Then it follows that there is a y = y 1 y 2 y n W n D such that d ˜ n x y ϵ . Also note that M D is a d ˜ n - n ϵ -separated subset of W n D with maximum cardinality; there is a z M D such that d ˜ n y z ϵ . Hence we have

d ˜ n x z d ˜ n x y + d ˜ n y z < 3 ϵ .

This yields the claim that M is a d ˜ n - n ϵ -spanning subset of O n . So we have r n 3 ϵ O n M , where M denotes the cardinality of M . Using the claim we have

r n 3 ϵ O n M Y max M D : W n D Y s n ϵ O n , P n s max , P n n ϵ .

Taking limits as the requirements of the related definitions of entropies establishes the desired inequality. This completes the proof.□

Let P δ be a finite cover of a compact metric space X consisting of open balls with radius less than some δ > 0 . Write F X = f i : f i : X X is continous i 1 and P X δ = P δ P δ .

Theorem 5.3.

h top F X = h p , P X δ F X = lim ϵ 0 lim δ 0 lim sup n 1 n log s max , P δ n ϵ .

Proof. Note that lim n 1 n log P δ = 0 . Then, by Theorem 5.2, we have the former equality. Now we show the later equality.

Clearly, s n ϵ O n s max , P X δ n ϵ for any δ > 0 , so we have

lim sup n log s n ϵ O n lim δ 0 lim sup n 1 n log s max , P X δ n ϵ .

This implies

h top F X lim ε 0 lim δ 0 limsup n 1 n log s max , P X δ n ε . E22

On the other hand, from the proof of Theorem 5.2, it follows that

r n 3 ϵ O n s n ϵ O n , P δ s max , P X δ n ϵ

for any integer n 1 , any sufficiently small ϵ > 0 and any δ > 0 . Noting that s n ϵ O n , P δ P δ for any integer n 1 , then we have

lim sup n 1 n log r n 3 ϵ O n lim δ 0 lim sup n 1 n log s max , P X δ n ϵ .

This implies

h top F X lim ϵ 0 lim δ 0 lim sup n 1 n log s max , P X δ n ϵ . E23

Thus, combining (22) and (23) gets the later equality. This completes the proof.□

Remark 5.4. The first equality of Theorem 5.3 is in fact a simpler version of Theorem 7.6 of [40] (a useful result for calculating the classical topological entropy) when restricting to the autonomous discrete dynamical systems.

Given a nonautonomous discrete dynamical system X , when does h top F = h p , P F for any cover P of X ? The following theorem gives an answer to this question.

Theorem 5.5. Let X F be a nonautonomous discrete dynamical system. Then h top F = h p , P F for any cover P of X if the following conditions hold:

(1) For each integer i 1 , there exists δ i > 0 such that d i + 1 f i x f i y d i x y whenever d i x y δ i for x , y X i .

(2) For each integer i 1 , every x X i + 1 has an open neighborhood U x whose preimage f i 1 U x is an union of disjoint open sets on each of which f i is a homeomorphism.

(3) lim sup n 1 n log N ϵ n X n = 0 for every monotonic decreasing sequence ϵ n with lim n ϵ = 0 , where each N ϵ n X n denotes the minimal cardinality of the open cover of X consisting of open ϵ n -ball for the compact metric space X n .

Proof. It suffices to show that h b , P F = 0 for any cover P of X by Theorem 5.2. Let P max = P i , max : i 1 be the cover of X in which each P i , max cover X i consisting to singletons of X i , i.e., P i , max = z : z X i . It is easy to see that h b , P F h b , P max F for any cover P of X . So from Theorem 5.2, it follows that what we want to prove is h b , P max F = 0 .

For each n 2 , by condition (1), there exists a δ n 1 > 0 such that

d n f n 1 x f n 1 y d n 1 x y

for any x , y X n 1 whenever d n 1 x y δ n 1 . Also, by condition (2) and the compactness of X n , there exists an ϵ n > 0 such that the ϵ n -ball B x n ϵ n about any point x n X n has preimage f n 1 1 B x n ϵ equals the union of disjoint open sets of diameter less than δ n 1 . Then we get a sequence ϵ n . Furthermore, we can take ϵ n such that ϵ n is monotonic decreasing sequence and lim n ϵ n = 0 .

Now, given y n X n and x ˜ = x 1 x 2 x n W n B y n ϵ n , we want to find a point y ˜ = y 1 y 2 y n O n with d ˜ n x ˜ y ˜ = d n x n y n and then d ˜ n x ˜ y ˜ < ϵ n . In fact, for 1 < k < n , we can easily find a point y k y n i = k n X i with d j x j y j ϵ j and d j + 1 x j + 1 y j + 1 d j x j y j , for j = n 1 , n 2 , , k . Let V be the piece of f k 1 1 B x k ϵ k with x k 1 V . Since y k B x k ϵ k , there is a unique point y k 1 V f k 1 1 y k such that d k 1 x k 1 y k 1 < δ k 1 . Then we have

d k 1 x k 1 y k 1 d k x k y k d n x n y n < ϵ n < ϵ k 1 .

This argument shows that r n ϵ n O n , \ , max N ϵ n X n . Thus, by condition (3), we get

lim sup n 1 n log r n ϵ n O n , \ , max lim sup n 1 n log N ϵ n X n = 0 .

For any sufficiently small ϵ > 0 , there exists N > 0 such that ϵ n < ϵ for any n N . Then we have r n ϵ O n , P n , max r n ϵ n O n , P n , max and hence h b , P max F = 0 . This completes the proof.□

## 6. Conclusion

Several important entropy-like invariants based on the preimage structure for non-invertible maps have been defined and studied by some authors. In this chapter, we first further study the preimage entropy for topological dynamical system from the view of localization. We show that the topological entropy for an open cover can be characterized by pseudo-orbits (Theorems 2.3, 4.2, and 4.3). We also establish an inequality relating the topological entropy for open covers and several preimage entropy invariants, which is viewed as the local version of the Hurley’s inequality (Theorem 3.1). Finally, we discuss the topological entropy for nonautonomous discrete dynamical systems by introducing two entropy-like invariants called the partial entropy and bundle-like entropy. We establish some relationships among such two invariants and the topological entropy (Theorem 5.2, 5.3, and 5.5).

## Acknowledgments

This work was carried out when Kesong Yan visited the Michigan State University. Kesong Yan sincerely appreciates the warm hospitality of Professor Huyi Hu. We thank the anonymous referees for their useful comments and helpful suggestions that improved the manuscript. The authors are supported by NNSF of China (11861010,11761012) and NSF for Distinguished Young Scholar of Guangxi Province (2018GXNSFFA281008). The first author is supported by the Cultivation Plan of Thousands of Young Backbone Teachers in Higher Education Institutions of Guangxi Province, Program for Innovative Team of Guangxi University of Finance and Economics, and Project of Guangxi Key Laboratory Cultivation Base of Cross-border E-commerce Intelligent Information Processing (201801ZZ03).

## References

1. 1. Hurley H. On topological entropy of maps. Ergodic Theory and Dynamical Systems. 1995;15:557-568. DOI: 10.1017/S014338570000852X
2. 2. Adler RL, Konheim AG, McAndrew MH. Topological entropy. Transactions of the American Mathematical Society. 1965;114:309-319. DOI: 10.1090/S0002-9947-1965-0175106-9
3. 3. Dinaburg EI. On the relations among various entropy characteristic of dynamical systems. Izvestiya Rossiiskoi Akademii Nauk. Seriya Matematicheskaya. 1971;35:324-366
4. 4. Bowen R. Entropy for group endomorphisms and homogeneous spaces. Transactions of the American Mathematical Society. 1971;153:401-414. DOI: 10.1090/S0002-9947-1971-0274707-X
5. 5. Langevin R, Walczak P. Entropie d’une dynamique [Entropy of a dynamic]. Comptes Rendus de I’ Académie des Sciences-Series I-Mathematics. 1991;312:141-144. (in French)
6. 6. Ghys E, Langevin R, Walczak P. Entroie geoetrique des feuilletages [Geometric entropy of foliations]. Acta Math. 1988;160:105-142 (in French). DOI: 10.1007/BF02392274
7. 7. Langevin R, Przytycki F. Entropie de I’image inverse d’une application. [Entropy of the inverse of a mapping]. Bulletin de la Société Mathématique de France. 1992;120:237-250. DOI: 10.24033/bsmf.2185 (in French)
8. 8. Nitecki Z, Przytycki F. Preimage entropy for mappings. International Journal of Bifurcation and Chaos. 1999;9:1815-1843. DOI: 10.1142/S0218127499001309
9. 9. Biś A. Entropies of a semigroup of maps. Discrete and Continuous Dynamical Systems. 2004;11:639-648. DOI: 10.3934/dcds.2004.11.639
10. 10. Nitecki Z. Topological entropy and the preimage structure of maps. Real Analysis Exchange. 2003;29:9-43. DOI: 10.14321/realanalexch.29.1.0009
11. 11. Yan K, Zeng F. Variational principles of partial pre-image entropy and conditional pre-image entropy. Scientia Sinica Mathematica. 2019;49:681-698. DOI: 10.1360/SCM-2016-0381. (in Chinese)
12. 12. Cheng W. Relations among conditional entropy, topological entropy and pointwise preimage entropy [thesis]. East Lansing: Michigan State University; 2004
13. 13. Cheng W. Pre-image entropy. Ergodic Theory and Dynamical Systems. 2005;25:1091-1113. DOI: 10.1017/S0143385704000240
14. 14. Cheng W. Forward generator for preimage entropy. Pacific Journal of Mathematics. 2006;223:5-16. DOI: 10.2140/pjm.2006.223.5
15. 15. Cheng W. Two-point pre-image entropy. Discrete and Continuous Dynamical Systems. 2007;17:107-119. DOI: 10.3934/dcds.2007.17.107
16. 16. Cheng W. On preimage entropy of a sequential mapping. Open Systems and Information Dynamics. 2008;15:345-357. DOI: 10.1142/S1230161208000237
17. 17. Cheng W. Pre-image entropy for free semigroup actions. Chaos, Solitons and Fractals. 2016;91:286-290. DOI: 10.1016/j.chaos.2016.06.011
18. 18. Fiebig D, Fiebig U, Nitecki Z. Entropy and preimage sets. Ergodic Theory and Dynamical Systems. 2003;23:1785-1806. DOI: 10.1017/S0143385703000221
19. 19. Yan K, Zeng F, Zhang G. Equilibrium states for pre-image pressure. Nonlinear Dynamics and Systems Theory. 2010;10:409-416
20. 20. Yan K, Zeng F, Zhang G. Differentiability properties of the pre-image pressure. Discrete Dynamics in Nature and Society; 2012. DOI: 10.1155/2012/951691, Article ID 951691, 14 pages
21. 21. Zeng F, Yan K, Zhang G. Pre-image pressure and invariant measures. Ergodic Theory and Dynamical Systems. 2007;27:1037-1052. DOI: 10.1017/S0143385706000812
22. 22. Zhu Y. Preimage entropy for random dynamical systems. Discrete and Continuous Dynamical Systems. 2007;18:829-851. DOI: 10.3934/dcds.2007.18.829
23. 23. Zhu Y, Li Z, Li X. Preimage pressure for random transformations. Ergodic Theory and Dynamical Systems. 2009;29:1669-1687. DOI: 10.1017/S0143385708000758
24. 24. Blanchard F. Fully positive topological entropy and topological mixing. In: Walters P, editor. Symbolic Dynamics and its Applications. Contemporary Mathematics. Vol. 135. Providence: RI: American Mathematical Society; 1992. pp. 95-105. DOI: 10.1090/conm/135
25. 25. Blanchard F, Lacroix Y. Zero-entropy factors of topological flows. Proceedings of the American Mathematical Society. 1993;119:985-992. DOI: 10.1090/S0002-9939-1993-1155593-2
26. 26. Blanchard F. A disjointness theorem involving topological entropy. Bulletin de la Société Mathématique de France. 1993;121:565-578. DOI: 10.24033/bsmf.2216
27. 27. Blanchard F, Host B, Maass A, Martinez S, Rudolph D. Entropy pairs for a measure. Ergodic Theory and Dynamical Systems. 1995;15:621-632. DOI: 10.1017/S0143385700008579
28. 28. Glasner E, Weiss B. Topological entropy of extensions. In: Petersen Karl E, editor. Ergodic Theory and its Connections with Harmonic Analysis. Cambridge: Cambridge University Press; 1995. pp. 299-307. DOI: 10.1017/CBO9780511574818.011
29. 29. Huang W, Ye X. A local variational relation and applications. Israel Journal of Mathematics. 2006;151:237-279. DOI: 10.1007/BF02777364
30. 30. Glasner E, Ye X. Local entropy theory. Ergodic Theory and Dynamical Systems. 2009;29:321-356. DOI: 10.1017/S0143385708080309
31. 31. Bowen R. Equilibrium States and the Ergodic Theory of Axiom Diffeomorphisms, Lecture Notes in Math. Second revised ed. Vol. 470. New York: Springer-Verlag, Berlin; 2008. 75 p. DOI: 10.1007/BFb0081284
32. 32. Conley C. Isolated Invariant Sets and the Morse Index. Vol. 38. Providence: CBMS Regional Conference Series in Mathematics, American Mathematical Society; 1978. 89 p. DOI: 10.1090/cbms/038
33. 33. Hammel SM, Yorke JA, Grebogi C. Do numerical orbits of chaotic dynamical processes represent true orbits? Journal of Complexity. 1987;3:136-145. DOI: 10.1016/0885-064X(87)90024-0
34. 34. Hammel SM, Yorke JA, Grebogi C. Numerical orbits of chaotic processes represent true orbits. Bulletin of the American Mathematical Society. 1988;19:465-469
35. 35. Barge M, Swanson R. Rotation shadowing properties of circle and annulus maps. Ergodic Theory and Dynamical Systems. 1988;8:509-521. DOI: 10.1017/S0143385700004661
36. 36. Misiurewicz M. Remark on the definition of topological entropy. In: Lara-Carrero L, Lewowicz J, editors. Dynamical Systems and Partial Differential Equations (Caracas, 1984). Caracas: Equinoccio; 1986. pp. 65-67
37. 37. Barge M, Swanson R. Pseudo-orbits and topological entropy. Proceedings of the American Mathematical Society. 1990;109:559-566. DOI: 10.1090/S0002-9939-1990-1012923-7
38. 38. Kolyada S, Snoha L. Topological entropy of nonautonomous dynamical systems. Random & Computational Dynamics. 1996;4:205-233
39. 39. Zeng F. A note on topological entropy of maps. Northeastern Mathematical Journal. 1997;13:250-255. DOI: 10.13447/j.1674-5647.1997.04.012
40. 40. Walters P. An introduction to ergodic theory. In: Graduate Texts in Mathematics. Vol. 79. New York: Berlin: Springer-Verlag; 1982. 256 p. DOI: 10.1007/978-1-4612-5775-2
41. 41. Kolyada S, Snoha L. On topological dynamics of sequences of continuous map. International Journal of Bifurcation and Chaos. 1995;5:1437-1438. DOI: 10.1142/S0218127495001125

Written By

Kesong Yan and Fanping Zeng

Submitted: May 14th, 2019 Reviewed: August 5th, 2019 Published: September 23rd, 2019