Chapter 11

Brownian motion as a random fractal

[13/100%]
Show
LaTeX
Problem 11.1

Verify that s - dimensional Hausdorff measure is an outer measure.

?
Problem 11.2

Show that the dd - dimensional Hausdorff measure of a bounded open set U⊂RdU \subset \mathbb {R}^{d} is finite and positive: 0<Hd(U)<∞0 < \mathscr {H}^{d}(U) < \infty. In particular, dim⁡U=d\operatorname {dim} U = d.

?
Problem 11.3

Let E⊂RdE \subset \mathbb {R}^{d}. Show that Hausdorff dimension dim⁡E\operatorname {dim} E coincides with the numbers

sup⁡{α⩾0:Hα(E)=∞},sup⁡{α⩾0:Hα(E)>0}inf⁡{α⩾0:Hα(E)<∞},inf⁡{α⩾0:Hα(E)=0}. \begin{array}{cll} \sup \left\{ \alpha \geqslant 0: \mathscr {H}^{\alpha }(E) = \infty \right\} , & \sup \left\{ \alpha \geqslant 0: \mathscr {H}^{\alpha }(E) > 0\right\} \\ \inf \left\{ \alpha \geqslant 0: \mathscr {H}^{\alpha }(E) < \infty \right\} , & \inf \left\{ \alpha \geqslant 0: \mathscr {H}^{\alpha }(E) = 0\right\} . \end{array}
?
(a)
(b)
Problem 11.4

Let (Ej)j⩾1\left(E_{j}\right)_{j \geqslant 1} be subsets of Rd\mathbb {R}^{d} and E:=⋃j⩾1EjE: = \bigcup_{j \geqslant 1} E_{j}. Show that dim⁡E=sup⁡j⩾1dim⁡Ej\operatorname {dim} E = \sup_{j \geqslant 1} \operatorname {dim} E_{j}.

?
Problem 11.5

Let E⊂RdE \subset \mathbb {R}^{d}. Show that dim⁡(E×Rn)=dim⁡E+n\operatorname {dim}\left(E \times \mathbb {R}^{n}\right) = \operatorname {dim} E+n. Does a similar formula hold for E×FE \times F where E⊂RdE \subset \mathbb {R}^{d} and F⊂RnF \subset \mathbb {R}^{n} ?

?
Problem 11.6

Let f:Rd→Rnf: \mathbb {R}^{d} \rightarrow \mathbb {R}^{n} be a bi - Lipschitz map, i.e. both ff and f−1f^{ - 1} are Lipschitz continuous. Show that dim⁡f(E)=dim⁡E\operatorname {dim} f(E) = \operatorname {dim} E. Is this also true for a Hölder continuous map with index γ∈(0,1)\gamma \in (0,1) ?

?
Problem 11.7

Let CC denote Cantor's discontinuum which is obtained if we remove recursively the open middle third of any remaining interval:

[0,1]⇝[0,13]∪[23,1]⇝[0,19]∪[29,13]∪[23,79]∪[89,1]⇝… [0,1] \leadsto \left[0, \frac{1}{3}\right] \cup \left[\frac{2}{3}, 1\right] \leadsto \left[0, \frac{1}{9}\right] \cup \left[\frac{2}{9}, \frac{1}{3}\right] \cup \left[\frac{2}{3}, \frac{7}{9}\right] \cup \left[\frac{8}{9}, 1\right] \leadsto \ldots

see e.g. [232, Problem 7.12] or [236, § 2.5]. Show that C=f1(C)∪f2(C)C = f_{1}(C) \cup f_{2}(C) with f1(x)=13xf_{1}(x) = \frac{1}{3} x and f2(x)=13x+23f_{2}(x) = \frac{1}{3} x+\frac{2}{3} and dim⁡C=log⁡2/log⁡3\operatorname {dim} C = \log 2 / \log 3.

?
Problem 11.8

Let (Bt)t⩾0\left(B_{t}\right)_{t \geqslant 0} be a BMd\mathrm{BM}^{d}. Show that for λ<d\lambda < d the expectation E[∣B1∣−λ]\mathbb {E}\left[\left|B_{1}\right|^{ - \lambda }\right] is finite.

?
Problem 11.9

Let (Bt)t⩾0\left(B_{t}\right)_{t \geqslant 0} be a one - dimensional Brownian motion. Show that

dim⁡B−1(A)⩽12+12dim⁡A a.s. for all A∈B(R) \operatorname {dim} B^{ - 1}(A) \leqslant \frac{1}{2}+\frac{1}{2} \operatorname {dim} A \quad \text{ a.s. for all } A \in \mathscr {B}(\mathbb {R})
?
Problem 11.10

Let F⊂RF \subset \mathbb {R} be a non - void perfect set, i.e. a closed set such that each point in FF is an accumulation point of FF. Show that a perfect set is uncountable.

?
Problem 11.11

Let (Bt)t⩾0\left(B_{t}\right)_{t \geqslant 0} be a BM1\mathrm{BM}^{1} and denote by gtg_{t} the last zero before time t>0t > 0.

?
(a)

Use Theorem 11.25 to give a further proof of Lévy's arc - sine law (cf. Theorem 6.20): P(gt<s)=2πarcsin⁡st\mathbb {P}\left(g_{t} < s\right) = \frac{2}{\pi } \arcsin \sqrt{\frac{s}{t}} for all s∈[0,t]s \in [0, t].

(b)

Find the conditional density P(∣Bt∣∈dy∣gt=s)\mathbb {P}\left(\left|B_{t}\right| \in d y \mid g_{t} = s\right).

(c)

Integrate the formula appearing in Theorem 11.25 to get

2πtexp⁡(−y22t)=∫0t2πsy2π(t−s)3exp⁡(−y22(t−s))ds \sqrt{\frac{2}{\pi t}} \exp \left( - \frac{y^{2}}{2 t}\right) = \int _{0}^{t} \sqrt{\frac{2}{\pi s}} \frac{y}{\sqrt{2 \pi (t - s)^{3}}} \exp \left( - \frac{y^{2}}{2(t - s)}\right) d s

and give a probabilistic interpretation of this identity.

Problem 11.12

Show that Corollary 11.26 also follows from Lemma 11.23.

?
Problem 11.13

Let (Bt)t⩾0,dt\left(B_{t}\right)_{t \geqslant 0}, d_{t} and gtg_{t} be as in Corollary 11.26. Define

Lt−:=t−gt and Lt:=dt−gt L_{t}^{ - }: = t - g_{t} \quad \text{ and } \quad L_{t}: = d_{t} - g_{t}

Find the laws of (Lt−,Lt),Lt−,Lt\left(L_{t}^{ - }, L_{t}\right), L_{t}^{ - }, L_{t} and LtL_{t} conditional on Lt−=rL_{t}^{ - } = r.

Find the probability that P(Lt>r+s∣Lt−=r)\mathbb {P}\left(L_{t} > r+s \mid L_{t}^{ - } = r\right), i.e. the probability that an excursion continues for ss units of time if we know that it is already rr units of time under way.

?