ABOUT CHAOS AND SENSITIVITY IN TOPOLOGICAL DYNAMICS

Similar documents
Lecture Notes for Analysis Class

McGill University Math 354: Honors Analysis 3 Fall 2012 Solutions to selected problems

Sequences and Series of Functions

lim za n n = z lim a n n.

(A sequence also can be thought of as the list of function values attained for a function f :ℵ X, where f (n) = x n for n 1.) x 1 x N +k x N +4 x 3

On groups of diffeomorphisms of the interval with finitely many fixed points II. Azer Akhmedov

Chapter 2. Periodic points of toral. automorphisms. 2.1 General introduction

REGULARIZATION OF CERTAIN DIVERGENT SERIES OF POLYNOMIALS

Singular Continuous Measures by Michael Pejic 5/14/10

Metric Space Properties

MAT1026 Calculus II Basic Convergence Tests for Series

Math 61CM - Solutions to homework 3

Chapter 6 Infinite Series

Math 220A Fall 2007 Homework #2. Will Garner A

Introduction to Optimization Techniques

HOMEWORK #4 - MA 504

Math Solutions to homework 6

If a subset E of R contains no open interval, is it of zero measure? For instance, is the set of irrationals in [0, 1] is of measure zero?

Ma 530 Introduction to Power Series

Chapter 1. Complex Numbers. Dr. Pulak Sahoo

Chapter 0. Review of set theory. 0.1 Sets

The Boolean Ring of Intervals

Convergence of random variables. (telegram style notes) P.J.C. Spreij

The Borel hierarchy classifies subsets of the reals by their topological complexity. Another approach is to classify them by size.

6. Uniform distribution mod 1

Axioms of Measure Theory

Theorem 3. A subset S of a topological space X is compact if and only if every open cover of S by open sets in X has a finite subcover.

n=1 a n is the sequence (s n ) n 1 n=1 a n converges to s. We write a n = s, n=1 n=1 a n

Properties of Fuzzy Length on Fuzzy Set

Notes #3 Sequences Limit Theorems Monotone and Subsequences Bolzano-WeierstraßTheorem Limsup & Liminf of Sequences Cauchy Sequences and Completeness

University of Colorado Denver Dept. Math. & Stat. Sciences Applied Analysis Preliminary Exam 13 January 2012, 10:00 am 2:00 pm. Good luck!

Dynamics of Piecewise Continuous Functions

Integrable Functions. { f n } is called a determining sequence for f. If f is integrable with respect to, then f d does exist as a finite real number

f n (x) f m (x) < ɛ/3 for all x A. By continuity of f n and f m we can find δ > 0 such that d(x, x 0 ) < δ implies that

PRELIM PROBLEM SOLUTIONS

Lecture 2. The Lovász Local Lemma

The log-behavior of n p(n) and n p(n)/n

Infinite Sequences and Series

6.3 Testing Series With Positive Terms

62. Power series Definition 16. (Power series) Given a sequence {c n }, the series. c n x n = c 0 + c 1 x + c 2 x 2 + c 3 x 3 +

MASSACHUSETTS INSTITUTE OF TECHNOLOGY 6.436J/15.085J Fall 2008 Lecture 19 11/17/2008 LAWS OF LARGE NUMBERS II THE STRONG LAW OF LARGE NUMBERS

Final Solutions. 1. (25pts) Define the following terms. Be as precise as you can.

Week 5-6: The Binomial Coefficients

Commutativity in Permutation Groups

Math F215: Induction April 7, 2013

Best bounds for dispersion of ratio block sequences for certain subsets of integers

Period Function of a Lienard Equation

M17 MAT25-21 HOMEWORK 5 SOLUTIONS

Discrete-Time Systems, LTI Systems, and Discrete-Time Convolution

2 Banach spaces and Hilbert spaces

Chapter 7 Isoperimetric problem

On Syndetically Hypercyclic Tuples

MATH 112: HOMEWORK 6 SOLUTIONS. Problem 1: Rudin, Chapter 3, Problem s k < s k < 2 + s k+1

FUNDAMENTALS OF REAL ANALYSIS by

A Proof of Birkhoff s Ergodic Theorem

A gentle introduction to Measure Theory

Analytic Continuation

Math 299 Supplement: Real Analysis Nov 2013

Weakly Connected Closed Geodetic Numbers of Graphs

Chapter 3. Strong convergence. 3.1 Definition of almost sure convergence

On Topologically Finite Spaces

Product measures, Tonelli s and Fubini s theorems For use in MAT3400/4400, autumn 2014 Nadia S. Larsen. Version of 13 October 2014.

Archimedes - numbers for counting, otherwise lengths, areas, etc. Kepler - geometry for planetary motion

2.1. The Algebraic and Order Properties of R Definition. A binary operation on a set F is a function B : F F! F.

Measure and Measurable Functions

Lecture XVI - Lifting of paths and homotopies

Research Article A Note on Ergodicity of Systems with the Asymptotic Average Shadowing Property

MATH4822E FOURIER ANALYSIS AND ITS APPLICATIONS

REAL ANALYSIS II: PROBLEM SET 1 - SOLUTIONS

TR/46 OCTOBER THE ZEROS OF PARTIAL SUMS OF A MACLAURIN EXPANSION A. TALBOT

Lecture 19: Convergence

1 Lecture 2: Sequence, Series and power series (8/14/2012)

1+x 1 + α+x. x = 2(α x2 ) 1+x

The average-shadowing property and topological ergodicity

MAS111 Convergence and Continuity

Math 341 Lecture #31 6.5: Power Series

1 Introduction. 1.1 Notation and Terminology

1 Convergence in Probability and the Weak Law of Large Numbers

Fall 2013 MTH431/531 Real analysis Section Notes

The Random Walk For Dummies

McGill University Math 354: Honors Analysis 3 Fall 2012 Solutions to selected problems. x y

Real Variables II Homework Set #5

3 Gauss map and continued fractions

Beurling Integers: Part 2

Exponential Functions and Taylor Series

On Random Line Segments in the Unit Square

Topological Folding of Locally Flat Banach Spaces

ON THE FUZZY METRIC SPACES

Ma 530 Infinite Series I

NBHM QUESTION 2007 Section 1 : Algebra Q1. Let G be a group of order n. Which of the following conditions imply that G is abelian?

Chain conditions. 1. Artinian and noetherian modules. ALGBOOK CHAINS 1.1

Solution. 1 Solutions of Homework 1. Sangchul Lee. October 27, Problem 1.1

CANTOR SETS WHICH ARE MINIMAL FOR QUASISYMMETRIC MAPS

Section 1.4. Power Series

Chapter 2 The Solution of Numerical Algebraic and Transcendental Equations

SOME SEQUENCE SPACES DEFINED BY ORLICZ FUNCTIONS

FUCHSIAN GROUPS AND COMPACT HYPERBOLIC SURFACES

f(w) w z =R z a 0 a n a nz n Liouville s theorem, we see that Q is constant, which implies that P is constant, which is a contradiction.

AN ARC-LIKE CONTINUUM THAT ADMITS A HOMEOMORPHISM WITH ENTROPY FOR ANY GIVEN VALUE

HOMEWORK #10 SOLUTIONS

Transcription:

ABOUT CHAOS AND SENSITIVITY IN TOPOLOGICAL DYNAMICS EDUARD KONTOROVICH Abstract. I this work we uify ad geeralize some results about chaos ad sesitivity. Date: March 1, 005. 1

1. Symbolic Dyamics Defiitio 1.1. The sequece space o two symbols is the set Σ = {(s 0, s 1, s,...) : s j = 0 or 1} Propositio 1.. The distace d o Σ is defied by d(s, t) = Σ s i t i i. Proof. Let s = (s 0, s 1, s,...), t = (t 0, t 1, t,...) ad u = (u 0, u 1, u,...). Clearly, d(s, t) 0 ad d(s, t) = 0 if ad oly if s = t. Sice s i t i = t i s i, it follows that d(s, t) = d(t, s). Fially, for ay three real umbers s i, t i, u i we have the usual triagle iequality from which we deduce that This completes the proof. s i t i + t i u i s i u i d(s, t) + d(t, u) d(s, u). Theorem 1.3. Let s, t Σ ad suppose s i = t i for i=0,1,,... The d(s, t) 1 Coversely, if d(s, t) < 1, the s i = t i for i. Proof. If s i = t i for i, the d(s, t) = Σ s i t i i=+1 i Σ 1 i=+1 i = 1 O the other had, if s i t i for some i, the we must have d(s, t) 1 1 i. Cosequetly if d(s, t) < 1, the s i = t i for i = 0, 1,,... Defiitio 1.4. The shift map σ : Σ Σ is defied by σ(s 0, s 1, s,...) = (s 1, s, s 3,...). Our first observatio about this map is that the subset of Σ that cosist of all periodic poits i Σ is a dese subset. To see why this is true, we must show that, give ay poit s = (s 0, s 1, s,...) i Σ, we ca fid a periodic poit arbitrarily close to s. So let ɛ > 0. Let s choose a iteger umber so that 1 < ɛ. We may ow write dow a explicit periodic poit withi 1 uits of s. Let t = (s 0, s 1,..., s, s 0, s 1,..., s ). The first + 1 etries of s ad t are the same. By Theorem 1.3 this meas that d(s, t ) 1 < ɛ. Clearly t is a periodic poit of period + 1 for σ : Σ Σ. Sice ɛ ad s were arbitrary, we have succeeded i fidig a periodic poit arbitrarily close to ay poit of Σ. Note that the sequece (of sequeces) {t } coverges to s i Σ as. A secod ad eve more iterestig property of σ is that there is a poit whose orbit is dese i Σ. That is we ca fid a orbit which comes arbitrarily close to ay poit of Σ. Clearly, this kid of orbit is far from periodic or evetually periodic. As above, we ca write dow such a orbit explicitly for σ. Cosider the poit ŝ = (0100011011000001...). I other words, ŝ is the sequece which cosists of all possible blocks of 0 s ad 1 s of legth, followed by all such blocks of legth, the legth 3, ad so forth. The poit ŝ has a orbit that forms a dese subset of Σ. To see this, we agai choose a arbitrary s = (s 0, s 1, s,...) Σ ad a ɛ > 0. Agai choose so that 1 < ɛ. Now we show that the orbit of ŝ comes withi 1 uits of s. Far to the right i the expressio for ŝ, there is a block of legth + 1 that cosists of the digits s 0 s 1...s. Suppose the etry s 0 is at the k-th.

place i the sequece. Now apply the shift map k times to ŝ. The the first +1 etries of σ k (ŝ) are precisely s 0 s 1...s. So by Theorem 1.3 we get d(σ k (ŝ), s) 1 < ɛ. There is a dyamical otio that is itimately related to the property of havig a dese orbit. This is the cocept of sesitivity. Defiitio 1.5. A metric dyamical system ((X, d), F ) depeds sesitively o iitial coditios, if there is a β > 0 such that for ay x X ad ay ɛ > 0 there is k N ad y X with d(x, y) < ɛ such that d(f k (x), F k (y) β 3 To see that shift map depeds sesitivity o iitial coditios, we choose β = 1. For ay s Σ ad ɛ > 0 oe ca agai pick N so that 1 < ɛ. Suppose t Σ satisfies d(s, t) < 1 but t s. The we kow that t i = s i for i = 0, 1,, 3,... However, sice t s there is k > such that s k t k. So s k t k = 1. Now cosider the sequece σ k (s) ad σ k (t). The iitial etries of each of these sequeces are differet, so we have d(σ k (s), σ k (t)) s k t k 0 + Σ 0 i = 1. This proves sesitivity for the shift.. Chaos by R.L.Devaey Defiitio.1. A dyamical system F : X X is chaotic if P 1 Periodic poits for F are dese. P F is trasitive. P 3 F depeds sesitivity o iitial coditios. Theorem.. The shift map σ : Σ Σ is a chaotic dyamical system. Theorem.3. The doublig map f is chaotic o the uit circle. Proof. Let S 1 be the uit circle {(x, y) : x + y = 1}. I the complex aalysis S 1 = {e iθ : θ R}. 1).Let e iθ S 1 ad U is a ope eighborhood of e iθ. Let A be a ope arc i U cotaiig e iθ, too. Note that f () (A) is a arc -th loger as A. There exists N such that f () (a) is a cover of S 1. Deote this iteratio by N. There are two poits x, y with d(f (N) (x), d(f (N) (y)) = 1. ). Let U, V are ope sets i S 1. If we precede as above, the for large eough we got that f () (U) covers S 1 ad therefore itersects V. 3). For poits of the form e iθ with period, the followig equatio holds e iθ. This meas that there periodic poits are uit roots with order 1. The set of all there poits is dese i S 1. Theorem.4. Let s suppose that g T = T f, ad T is cotiuous ad subjective map. If f is trasitive or periodic o X, the g is also trasitive or periodic o Y. Theorem.5. The fuctio g(x) = x chaotic order the iterval [, ]

4 Proof. Let T : S 1 [, ] be defied as T (e iθ ) = cosθ. It s clear to see that T is cotiuous ad subjective ad g T (e iθ ) = T f(e iθ ) = cosθ. f is chaotic o S 1, the by theorem.5, the fuctio g is chaotic too. The followig theorem which relates to chaos defiitio was published i 199. Theorem.6. Let (X, d) be a metric space that is iclude a ifiite set of poits. If the mappig f : X Y is cotiuous ad trasitive ad if a set of periodic poits is dese i X, the f is sesitive depedet o iitial coditios. Proof. Let choose two periodic poits q 1, q such that O(q 1 ) O(q ) =. Let δ 0 = d(o(q 1 ), O(q )). We ll show that the sesitive depedet o the iitial coditios holds whe δ = δ 0 8. Notice that δ 0 > 0, ad for every x X, other d(x, O(q 1 )) > δ 0 or d(x, O(q )) > δ 0. Let x X ad U be a ope set that icludes x. Let B δ (x) be a ope sphere with radius δ ad ceter x. Let p be a periodic poit i W = U B δ (x) with period. From this we coclude that oe of the poits q 1, q (deoted by q) has a orbit, for which d(x, O(q)) > 4δ. Let s defie V = i=0f ( i) (B δ (f (i) (q))). The set V is o empty, because q V ad V is ope. From the trasitivity of f exists a poit y W ad iteger umber k such that f (k) (y) V. Let j be a iteger part of k + 1. Cosequetly, k + 1 = j + r, whe r is the rest, 0 r < 1. Clearly, j k = r. It follows that 0 j k. By costructio, Let f (j) (y) = f (j k) (f (k) ) f (j k) (V ) B δ (f (j k) (q)). a = f (j) (y), b = f (j k) (q). Note that d(a, b) < δ. Let us use the triagle iequality for poits p, a, b ad x, p, b: Whe or By costructio or d(p, b) d(p, a) + d(a, b), d(x, b) d(x, p) + d(p, b). d(x, b) d(x, p) + d(p, a) + d(a, b), d(p, a) d(x, b) d(x, p) d(a, b). d(x, b) = d(x, f (j k) (q)) d(x, O(q)) 4δ. Sice p B δ (x), the d(x, p) < δ. From this it follows that d(a, b) > 4δ δ δ, d(f (j) (p), f (j) (y)) > δ.

Applyig the triagle iequality to the followig poits f (j) (x), f (j) (p), f (j) (y), we get that: d(f (j) (x), f (j) (p)) > δ or d(f (j) (x), f (j) (y)) > δ. 3. Topological trasitivity, miimality 5 Give a dyamical system (X, S) we let sx deote the image of x X uder the homeomorphism correspodig to the elemet s S. Let O S x be the orbit of x; i.e. the set {sx : s S}. O S x will deote the orbit closure of x. If (X, S)is a system ad Y a closed S-ivariat subset, the we say that (Y, S), the restricted actio, a subsystem of (X, S). Defiitio 3.1. The dyamical system (X, S) is called topologically trasitive or just trasitive, if for every pair of o-empty ope sets U, V i X there exists s S with su V. Defiitio 3.. The dyamical system (X, S) us called poit trasitive, if there exists poit x 0 X with O S x = X. Such x 0 is called a trasitive poit. Example 3.3. If X = S 1 ad S = {f () : N, f(z) = z }, the this dyamical system is trasitive. Defiitio 3.4. We dyamical system (X, S) is called miimal, if O S x = X for every x X. Defiitio 3.5. A poit x i dyamical system (X, S) is called miimal(or almost periodic), if the subsystem O S x is miimal. Defiitio 3.6. If set of miimal poits is dese i X, we say that (X, S) satisfies the Brostei coditio. If, i additio, the system (X, S) is trasitive, we say that it is a M-system. Defiitio 3.7. A poit x X is a periodic poit, if O S x is fiite set. If (X,S) is a trasitive system ad set of periodic poit is dese i X, the we say that it is a P - system. For a system (X, S) ad subsets A, X X, we use the followig atatio N(A, B) = {s S : sa B }. I particular, for A = {x} we write N(x, B) = {s S : sx B}. Defiitio 3.8. A subset P S is (left) sydetic, if there exists a fiite set F S such that F P = S. Theorem 3.9. The followig are equivalet: 1 (X, S) is miimal. (X, S) is a trasitive ad for every x X ad eighborhood U of x, the set N(x, U) = {s S : sx B} is sydetic i S. Proof. (a) = (b): If (X, S) is a miimal system the for every o-empty set U i X there exists a fiite subset F = {s 1, s,..., s k } i S with s i U = X. The j N(x, s i U) := S. But N(x, s j U) = s j N(x, U), the N(x, U) us sydetic. (a) = (b): For every x X ad every eighborhood U a set s j U is covered X. The for all ope set V exists s i S such that s i U V. Because X is a metric space we get that (X, S) is a miimal system.

6 4. Equicotiuity ad almost equicotiuity Defiitio 4.1. The system (X, S) is equicotiuous if the semigroup S acts equicotiuously o X; for every ɛ > 0 there exists δ > 0 such that d(x 1, x ) < δ implies d(sx 1, sx ) < ɛ, for every s S. Example 4.. Every isometric system is equicotiuous. It s clear to see that we must take ɛ = δ. Defiitio 4.3. Let (X, S) be a dyamical system. A poit x 0 X is called a equicotiuity poit if for every ɛ > 0 there exists a eighborhood U of x 0 such that for every y U ad every s S, d(sx 0, sy) ɛ. Propositio 4.4. A system (X,S) is equicotiuous iff every x X is a equicotiuity poit. Proof. : It s clearly. : Give ɛ > 0, let I := {U x : x X} be a collectio of eighborhoods as i the defiitio of equicotiuity poits. Ay Lebesgue umber δ for the ope cover I will serve for the equicotiuity coditio. Defiitio 4.5. The dyamical system (X, S) is called almost equicotiuous (or is a AE-system) if the subset EQ(X) of equicotiuity poits is a dese subset of X. Propositio 4.6. A miimal almost equicotiuous system is equicotiuous. Proof. Let x 0 be a trasitive poit ad x EQ(X). We shall show that also x 0 EQ(X). Give ɛ > 0 there exists δ > 0 such that for all x,, x,, B δ (x), d(x,, x,, ) < ɛ for all s S. Sice x 0 is a trasitive poit, there exists s, S ad η > 0 such that s, B η (x 0 ) B δ (x). Thus for every z B η (x 0 ) ad every s S we have d(ss, z, ss, x 0 ) < ɛ foe all s S; i.e. x 0 EQ(X) ad we coclude that the set of trasitive poits is cotaied i EQ(X). The a miimal almost equicotiuous system is equicotiuous. 5. Sesitive dyamical system Defiitio 5.1. We shell say that a system (X, S) is sesitive if it satisfies the followig coditio(sesitive depedece o iitial coditio): there exists a ɛ > 0 such that for all x X ad all δ > 0 there are some y B δ (x) ad s S with d(sx, sy) > ɛ. We say that (X, S) is o-sesitive otherwise. Propositio 5.. A trasitive dyamical system is almost equicotiuous iff it is osesitive. Proof. Clearly a almost equicotiuous system is o-sesitive. Coversely, beig osesitive meas that for every ɛ > 0 there exists x ɛ X ad δ ɛ > 0 such that for all y B δɛ (x ɛ ) ad every s S, d(sx ɛ, sy) < ɛ. For m N set V 1/m = B δ1/m (x 1/m ), U m = SV 1/m ad let R = m N U m. Suppose x R ad ɛ > 0. Choose m so that /m < ɛ, the x U m implies the exists s 0 S ad x 0 V 1/m such that s 0 x 0 = x. Put V = s 0 V 1/m. We ow see that for all y V ad every s S d(sx, sy) = d(ss 0 x 0, ss 0 y 0 ) < /m < ɛ, y 0 V 1/m. Thus the dese G δ set R cosists of equicotiuity poits. I this propositio we have see that miimality ad almost equicotiuity imply equicotiuity. We easily get a stroger result.

Theorem 5.3. A almost equicotiuous M-system (X, S) is miimal ad equicotiuous. Thus a M-system (hece also P -system) which is ot miimal equicotiuous is sesitive. Proof. Every trasitive poit is a equicotiuity poit. Let x 0 X be a equicotiuity ad trasitive poit. Give ɛ > 0 there exists a 0 < δ < ɛ such that x B δ (x 0 ) implies d(sx 0, sx) < ɛ for every s S. Let x, B δ (x 0 ) be a miimal poit. It that follows that T = {s S : d(sx 0, sx) ɛ} is a sydetic subset of S. Collectig these estimatios we get, for every s S, d(sx 0, x 0 ) d(sx 0, sx, ) + d(sx,, x 0 ) ɛ. Thus for each ɛ > 0 the set N(x 0, B ɛ (x 0 ) is a sydetic, whece x 0 is a miimal. It follows that X is a miimal, hace also equicotiuous by Propositio 4.6. Now, by propositio 5. we have that dyamical system which ot miimal or equicotiuous is sesitive. 7

8 Refereces 1. R.M. Crowover, Itroductio to Fractals ad Chaos, Joes ad Barlett Publishers, 1999.. R.L. Devaey, A first course i chaotic dyamical systems. Theory ad experimet, Joes ad Barlett Publishers, 199. 3. R.M Crowover, Itroductio to Fractals ad Chaos, Joes ad Barlett Publishers, 1999. 4. J.Baks, J.Brooks G.Gairs,G.Davis,P Stacey, O Devaey s defiitio of chaos,amer.math.mothly 99, 1993. 5. E. Glaser, Joiigs ad Ergodyc Theory, 00. Departmet of Mathematics, Bar-Ila Uiversity, 5900 Ramat-Ga, Israel E-mail address: bizia@walla.co.il URL: http://www.math.biu.ac.il/ katore