I am looking for the Fourier transform of the function $$ f(x):=e^{-i\pi x^2}, $$ where we define $$ \hat f(\xi):=\int_{\mathbb R}f(x)e^{-2\pi i x\xi}dx. $$ First of all, the function $f$ is not integrable, but it is bounded and hence it can be regarded as a tempered distribution on $\mathcal S$, the Schwartz function space.

Completing squares as usual, it strongly suggests that its Fourier transform, in the sense of tempered distributions, is given by $$ \frac 1 {\sqrt i} e^{ i\pi \xi^2}. $$ The only problem is that I dont know how to determine the branch of $\sqrt i$. Any suggestions?

| cite | improve this question | | | | |

$$\begin{aligned} \int_{-\infty}^\infty e^{-i\pi x^2} e^{-2\pi i x \xi} dx &= \int_{-\infty}^\infty e^{-i\pi [(x+\xi)^2-\xi^2]} dx \\ &= e^{i\pi \xi^2} \int_{-\infty}^\infty e^{-i\pi x^2} dx \\ &= \frac{e^{i\pi \xi^2}}{\sqrt{2}} \end{aligned}$$

The last integral (Fresnel function) is easily evaluated using a circular arc contour subtending an angle of $\pi/4$.

| cite | improve this answer | | | | |
  • $\begingroup$ Thank you so much! Then it seems that my guess for the answer was wrong, and we cannot treat $i$ as if it were real, right? $\endgroup$ – Singfook Sangwood Nov 8 '17 at 6:59
  • 2
    $\begingroup$ @ThomasYang The formula $$\int_{-\infty}^{\infty} e^{-a\pi x^2}e^{-2\pi i x \xi} dx = \frac{1}{\sqrt{a}} e^{-\pi \xi^2/a}$$ only holds when $a>0$. If you treat $a$ as a complex number, the path of integration will no longer be the real line, this causes problems. $\endgroup$ – pisco Nov 8 '17 at 7:06

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.