2
$\begingroup$

I would like to know the fourier transform of the product of the Cauchy probability distribution $f(x)=\frac{1}{\pi (1+x^2)}, -\infty<x<\infty$ with itself.

I know that the fourier transform of $f$ is $e^{-|t|}$. So the fourier transform of $f^2$ is the convolution of $e^{-|t|}$ with itself.

The convolution is $\int_{-\infty}^{\infty}e^{-|y|}\cdot e^{-|t-y|}dy$.

Can someone help me in evaluating this integral? Or is there any other way to find the fourier transform of $f^2$?

$\endgroup$
2
$\begingroup$

Convolution is fine: for evaluating $$\hat{f^2}(t)= \int_{-\infty}^{\infty}e^{-|y|} e^{-|t-y|}dy$$ split the integral up in parts. For $t>0$, we have $$\begin{align}\hat{f^2}(t)&= \left(\int_{-\infty}^0 + \int_0^t + \int_t^\infty\right) e^{-|y| -|t-y|}dy =\int_{-\infty}^0 e^{y-(t-y)}dy + \int_0^t e^{-y-(t-y)} dy+ \int_t^\infty e^{-y-(y-t)}dy \\ &= \frac{e^{-t}}2 + t e^{-t} +\frac{e^{-t}}2 = (t+1)e^{-t} \end{align}.$$ A similar calculation for $t<0$ yields the final result $$\hat{f^2}(t)=(|t|+1)e^{-|t|}.$$

Alternatively, you could directly arrive at the final result by calculating the Fourier integral using, e.g., the residue theorem.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.