Let $\alpha \neq 1.$
If $X,Y$ are two independent random variable such that $U=X+Y$ and $V=X+\alpha Y$ are independent, then $X$ and $Y$ are normally distributed.
In term of characteristic functions this means that $$\forall x, y \in \mathbb{R}, \phi_X(x+y)\phi_Y(x+\alpha y)=\phi_X(x)\phi_Y(x)\phi_X(y)\phi_Y(\alpha y) \ \ \ \ \ \ \ \ (E)$$
The particular case where $\alpha=-1$ was treated here.
Trying to use that way for the general case didn't work.
The main problem is how to solve the functional equation $(E).$
Is there any known way to solve the equation analytically? Is it possible to find a functional equation $(E_1)$ which only depends of $\phi_X$ and $(E_2)$ which only depends of $\phi_Y$?
(We should find $|\phi_X(x)|=e^{x^2c_1},|\phi_Y(x)|=e^{c_2x^2},c_1,c_2 \leq 0$)