Help with Inverse Fourier Transform Integral

eloso
Messages
2
Reaction score
0
Hi, I am taking a random process class and I came across a problem that has stumped me. I believe I know the end result but I would like to know how it is solved. I have been out of college for a while and I am a little rusty with integration.

Homework Statement



What I need is to find out the inverse Fourier transform of S(ω)=1/((1+ω^2)^2)

Homework Equations



I know that F^-1(S(ω)) = 1/(2∏) * ∫(S(ω)*e^(j*ω*t), ω,-∞,∞) where j is an imaginary number.

→ F^-1(S(ω)) = ∫(1/((1+ω^2)^2)*e^(j*ω*t), ω,-∞,∞)

The Attempt at a Solution



I tried to solve it using by parts, but it got very messy and I may be heading the wrong path. This is what I did:

u = (1+ω^2)^-2 du = -4*ω*(1+ω^2)^-3
dv = e^(j*ω*t) v = e^(j*ω*t)/(j*t)


I plugged it into the by parts equation and I felt like going on wouldn't help. I would appreciate any help. Thanks.
 
Last edited:
Physics news on Phys.org
eloso said:
Hi, I am taking a random process class and I came across a problem that has stumped me. I believe I know the end result but I would like to know how it is solved. I have been out of college for a while and I am a little rusty with integration.

Homework Statement



What I need is to find out the inverse Fourier transform of S(ω)=1/((1+ω^2)^2)


Homework Equations



I know that F^-1(S(ω)) = 1/(2∏) * ∫(S(ω)*e^(j*ω*t), ω,-∞,∞) where j is an imaginary number.

→ F^-1(S(ω)) = ∫(1/((1+ω^2)^2)*e^(j*ω*t), ω,-∞,∞)

The Attempt at a Solution



I tried to solve it using by parts, but it got very messy and I may be heading the wrong path. This is what I did:

u = (1+ω^2)^-2 du = -4*ω*(1+ω^2)^-3
dv = e^(j*ω*t) v = e^(j*ω*t)/(j*t)


I plugged it into the by parts equation and I felt like going on wouldn't help. I would appreciate any help. Thanks.

Mmm. Elementary integration techniques like integration by parts are going to get you nowhere here. You need to use complex analysis for integrals like this (which is pretty typical for Fourier integrals). (1+w^2)=(1+jw)(1-jw). So you have second order poles at +j and -j. You need to use a contour integral and the residue theorem. Any experience with that??
 
Unfortunately I don't, but I can definitely research the theorems. I appreciate the response.
 
There are two things I don't understand about this problem. First, when finding the nth root of a number, there should in theory be n solutions. However, the formula produces n+1 roots. Here is how. The first root is simply ##\left(r\right)^{\left(\frac{1}{n}\right)}##. Then you multiply this first root by n additional expressions given by the formula, as you go through k=0,1,...n-1. So you end up with n+1 roots, which cannot be correct. Let me illustrate what I mean. For this...

Similar threads

Replies
3
Views
2K
Replies
3
Views
1K
Replies
6
Views
3K
Replies
5
Views
1K
Replies
2
Views
2K
Replies
2
Views
7K
Replies
3
Views
1K
Replies
2
Views
2K
Back
Top