Checking the well-definedness of an adjoint operator

232 Views Asked by At

Let $T: L_2(\mathbb{R}) \to L_2(\mathbb{R})$ be the linear operator given by $$(Tx)(t) = \exp(it) x(t)$$ where $x \in L_2(\mathbb{R})$ and $t \in \mathbb{R}$.

Now I need to determine if the adjoint operator of $T$ is well-defined on the Hilbert space $L_2(\mathbb{R})$.

I am not sure what well-defined means in this context. One can easily check that $Tx \in L_2(\mathbb{R})$. I think in this context it just means that I should check if $D(T^*) = L_2(\mathbb{R})$ where

$$D(T^*) = \{y \in L_2(\mathbb{R}) | f_y: x \mapsto (Tx,y) \text{ is a bounded linear functional on } L_2(\mathbb{R})\}$$

is the domain of the adjoint operator. If that is true, we shall check if for each $y \in L_2(\mathbb{R})$ the (obviously linear) functional $f_y$ is bounded.

Let $y \in L_2(\mathbb{R})$, i.e. $\int_{\mathbb{R}} |y(t)|^2 dt < \infty$. It is $$ f_y(x) = (Tx,y) = \int_{\mathbb{R}} \exp(it) x(t) \overline{y(t)} dt. $$

Now I am stuck here and do not know what to do. I have the feeling that one has to choose a specific element $x \in L_2(\mathbb{R})$ depending on $y$ to deduce some information about the boundedness or maybe about $y$. But I have no idea how exactly this should work. The integral and the exponential function there is confusing me too.

Could someone please help me with this problem? Thanks!

1

There are 1 best solutions below

2
On BEST ANSWER
  1. You're absolutely correct, the domain of $T^\ast$ is $$D(T^\ast) := \{y \in L_2(\mathbb{R}) \mid \text{$(T(\cdot),y) : L_2(\mathbb{R}) \to \mathbb{C}$ is bounded}\}$$ by definition, and your task is to show that $D(T^\ast) = L_2(\mathbb{R})$.

  2. Consider $$(Tx,y) = \int_{-\infty}^\infty e^{it}x(t)\overline{y(t)}\,dt = \int_{-\infty}^\infty x(t) \overline{e^{-it}y(t)} \, dt.$$ Is Cauchy–Schwarz still applicable, and if so, what does it now tell you?