Paper 4, Section II, K

Probability and Measure
Part II, 2013

State Birkhoff's almost-everywhere ergodic theorem.

Let (Xn:nN)\left(X_{n}: n \in \mathbb{N}\right) be a sequence of independent random variables such that

P(Xn=0)=P(Xn=1)=1/2\mathbb{P}\left(X_{n}=0\right)=\mathbb{P}\left(X_{n}=1\right)=1 / 2

Define for kNk \in \mathbb{N}

Yk=n=1Xk+n1/2nY_{k}=\sum_{n=1}^{\infty} X_{k+n-1} / 2^{n}

What is the distribution of Yk?Y_{k} ? \quad Show that the random variables Y1Y_{1} and Y2Y_{2} are not independent.

Set Sn=Y1++YnS_{n}=Y_{1}+\cdots+Y_{n}. Show that Sn/nS_{n} / n converges as nn \rightarrow \infty almost surely and determine the limit. [You may use without proof any standard theorem provided you state it clearly.]