A question about convergence to infinity of a sequence of integrals of functions with complex values

41 Views Asked by At

I have a fixed real number $t\neq 0$ and a function with complex values but defined in $\mathbb{R}$: $$\ell: \mathbb{R} \to \mathbb{C}, \quad \ell(x) = e^{itx} - 1$$ For each $n \in \mathbb{N}$, I have a measure $M_n$ defined on the borelians of $\mathbb{R}$, such that: $$M_n \Big( \left[-\delta,\delta\right] \Big) = k_n, \quad k_n \uparrow\infty\,\, (n \to \infty)$$ for some $\delta>0$ fixed.

I want to show that $$\int_{-\delta}^{\delta} \ell(x) M_n d(x)\to \infty \quad (n \to \infty)$$

I'm trying to limit this integral to a lower bound that converges to infinity. For this, I'm trying to use $k_n$, but I don't know how to do that.

Help!

Update:

In fact, $M_n(\mathbb{R}) \to \infty$ with $M_n$ being a Lévy measure:

  1. $\forall \delta > 0,\,\, \exists \, C_\delta< \infty, \,\, M_n(\mathbb{R} \setminus B_{\delta} (0) ) \leq C_\delta,\,\, \forall n$ and

  2. $\exists \, C_0 < \infty, \, \int_{ |x| \leq 1} x^2 M_n (dx)< C_0,\,\, \forall n.$.

Futhermore, $M_n(E) = \sum_{j=1}^n \mu_{nj}(E)$ where $\mu_{nj}$ is not a dirac measure. This is in the context of the generalization of the central limit theorem for array of independent random variable!

1

There are 1 best solutions below

3
On

As stated this is false. (Perhaps some crucial hypothesis is missing). Let $M_n=n\delta_0$. The $M_n[-\delta,\delta]=n \to \infty$ but $\int l(x)dM_n(x)=0$ for all $n$.