Calculate the characteristic function of the sequence of random variables

395 Views Asked by At

I'm incredibly stuck on this problem.

Calculate the characteristic function (chf) of the sequence of random variables $$Y_n = \frac{(1)}{\sqrt n} \sum_{i=1}^n X_i$$ where $X_i \sim X$ and are independent. Also take the limit as n goes to positive infinity of the chf.

Got a hint to remember that if X and Y are independent, then

$$E(f(X)g(Y)) = E(f(X))E(g(Y))$$

To be honest, I'm completely lost on this. I know that the characteristic function of X for a given value u is $E_x(e^{iuX})$. Thanks!

1

There are 1 best solutions below

2
On BEST ANSWER

$E(e^{iuY_n})=E(e^{iu\frac{1}{\sqrt{n}}\sum_1^n{X_i}})=\prod_1^n{E(e^{iu\frac{X_i}{\sqrt{n}}})}=(E(e^{iu\frac{X}{\sqrt{n}}}))^n$

note if $E(X)\neq 0,E(Y)\rightarrow\infty $. Assume $E(X)=0$

Also for generic chf, we have $\psi(u)=E(e^{iuX})=1+\psi'(0)u+\frac{1}{2}\psi''(0)u^2+o(u^2)$

following the properties of chf, $\psi'(0)=iE(X)=0, \psi''(0)=i^2E(X^2)=-\sigma^2$

$(\psi(\frac{u}{\sqrt{n}}))^n\sim (1-\frac{1}{2}\sigma^2\frac{u^2}{n}+o(\frac{u^2}{n}))^n\rightarrow e^{-\frac{1}{2}\sigma^2 u^2}$