I read some proofs of Fourier transform of Heaviside function on this site but I don't really understand (because I haven't learned about distribution yet). I'm trying to derive it myself based on what I learned so far.

I use the following definitions of Heaviside and Dirac delta function:

$$H(t)= \lim_{a\to 0^{+}} H_a(t) \text {, where } H_a(t) = \begin{cases} e^{-at}, & t \geq 0 \\ 0, & \text{otherwise} \end{cases}$$ $$\delta(v)= \lim_{a\to 0^{+}} \delta_a(v) \text {, where } \delta_a(v)= \begin{cases} \frac {1}{a}, & |v| \leq \frac{a}{2} \\ 0, & \text{otherwise} \end{cases}$$ Now, we have: $$\mathcal {F}\{H(t)\}=\lim_{a\to 0^{+}}\mathcal {F}\{H_a(t)\}=\lim_{a\to 0^{+}} \int_{0}^{\infty} e^{-at}e^{-2i \pi vt}dt = \lim_{a\to 0^{+}} \frac {1}{a+2i \pi v}$$ $$=\lim_{a\to 0^{+}} \frac {a-2i \pi v}{a^2+4 \pi^2 v^2} =\lim_{a\to 0^{+}} \frac {a}{a^2+4 \pi^2 v^2}+\lim_{a\to 0^{+}} \frac {-2i \pi v}{a^2+4 \pi^2 v^2}$$

The second term corresponds to $\frac {1}{2i \pi v}$. But why does the first term correspond to $\frac {1}{2} \delta(v)$?

  • $\begingroup$ The first important thing about distributions is that they are not functions. They only 'live' under intervals. The definition for your $\delta$ doesn't properly reflect this. Which kind of limit are you using in the definition of your functions? What does the equality for the $\delta$ mean? Why can you exchange the limit and the integration? $\endgroup$ – Roland Dec 13 '15 at 20:07
  • $\begingroup$ I've just edited the definitions. The $\delta(v)$ is equal $+\infty$ if $v=0$, and $0$, otherwise, so I think it's correct by defining as above. I didn't exchange the limit and the integration, I calculated the integral first, then plugged the limit in. $\endgroup$ – SiXUlm Dec 13 '15 at 20:21
  • $\begingroup$ Ah I see what you mean by exchanging the limit and the integration. You are correct. I haven't found the reason. $\endgroup$ – SiXUlm Dec 13 '15 at 20:24
  • $\begingroup$ If you regard $\delta$ as a function, it's almost everywhere zero. Thus, every integral containing it will be zero. This is not the behaviour you're expecting from the $\delta$ functional. $\endgroup$ – Roland Dec 13 '15 at 23:15

The second term is one form of a nascent delta function. It is in fact the Poisson kernel in the upper half plane, which converges to the Dirac delta impulse:

$$\delta(x)=\lim_{a\rightarrow 0^+}\frac{1}{\pi}\frac{a}{a^2+x^2}$$


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.