1
$\begingroup$

I have a fourier transform which is

$$X(jω)=\frac{\cos(2ω)}{ω^2+ω+1}$$

and I am trying to calculate the value of the integral:

$$∫x(t)dt \ \ \ \ \ \ x \in (-\infty, \infty)$$.

I was thinking I could use Parseval's theorem and tried doing the integral of

$$\frac{1}{2\pi} \int \biggr |\frac{\cos(2ω)}{(ω^2+ω+1)}\biggr|^{2} d\omega$$

which comes out to roughly $1.789$ but after that I'm stuck. Would greatly appreciate any advice on how to proceed or if there is an easier way of doing this?

$\endgroup$
3
  • $\begingroup$ The problem I am working says X(jω). In the integral I solved the ^2 should be for the entire expression though. $\endgroup$
    – George12
    Commented Feb 12, 2015 at 14:43
  • $\begingroup$ So your complex variable is $j$, not $i$? $\endgroup$ Commented Feb 12, 2015 at 14:46
  • $\begingroup$ Seems that way, I don't know why. $\endgroup$
    – George12
    Commented Feb 12, 2015 at 14:47

2 Answers 2

1
$\begingroup$

Your integral, if it has a finite value and $X$ is continuous at $ω=0$, as it obviously is, has the value $\sqrt{2\pi}X(0)$.


This follows from the Fourier pair being $$ X(jω)=\frac1{\sqrt{2\pi}}\int_{\mathbb R}x(t)e^{-jωt}\,dt\text{ and }x(t)=\frac1{\sqrt{2\pi}}\int_{\mathbb R}X(jω)e^{jωt}\,dω $$ as long as the integrals can be evaluated (see extensions of Fourier transform to tempered distributions).

The first formula evaluated at $ω=0$ results in $$ X(j0)=\frac1{\sqrt{2\pi}}\int_{\mathbb R}x(t)\,dt. $$ There is no need to evaluate any norm integrals.

$\endgroup$
3
  • $\begingroup$ Thank you yes that seems to follow from my integral but the problem is that I have no clue what X(0) is and I am looking for a numeric value. $\endgroup$
    – George12
    Commented Feb 12, 2015 at 14:49
  • $\begingroup$ But your first formula gives you a very easy way to compute that value. $X(0)=X(j\cdot0)=\cos(0)/(1+0+0)$. $\endgroup$ Commented Feb 12, 2015 at 14:51
  • $\begingroup$ Ah right sure then we have sqrt(2pi)*1 but that gives me the integral of abs(x(t))^2 dt right? and I just want the integral of x(t). $\endgroup$
    – George12
    Commented Feb 12, 2015 at 14:56
0
$\begingroup$

You want the value of $$ \begin{align} &\frac{1}{2\pi}\int_{-\infty}^{\infty}\frac{\cos^{2}(2\omega)}{(\omega^{2}+\omega+1)^{2}}d\omega \\ & = \frac{1}{2\pi}\int_{-\infty}^{\infty}\frac{\cos^{2}(2\omega)}{(w-e^{2\pi i/3})^{2}(w-e^{-2\pi i/3})^{2}}d\omega \\ & = \frac{1}{4\pi}\int_{-\infty}^{\infty}\frac{\cos(4\omega)+1}{(w-e^{2\pi i/3})^{2}(w-e^{-2\pi i/3})^{2}}d\omega \\ & = \Re \frac{1}{4\pi}\int_{-\infty}^{\infty}\frac{e^{4i\omega}+1}{(w-e^{2\pi i/3})^{2}(w-e^{-2\pi i/3})^{2}}d\omega. \end{align} $$ Now you can close the integral in the upper half plane, which gives a value of $$ \left.\Re\left(\frac{2\pi i}{4\pi}\frac{d}{d\omega}\frac{e^{4i\omega}+1}{(w-e^{-2\pi i/3})^{2}}\right)\right|_{\omega=e^{2\pi i/3}}. $$

$\endgroup$

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .