Fourier transform: Difference between revisions

From Class Wiki
Jump to navigation Jump to search
Guenan (talk | contribs)
Guenan (talk | contribs)
Line 88: Line 88:
\mathcal{F}^{-1}[\Beta(f)]=\beta(t)=\int_{-\infty}^\infty \Beta(f) e^{j2\pi fkt}\,df
\mathcal{F}^{-1}[\Beta(f)]=\beta(t)=\int_{-\infty}^\infty \Beta(f) e^{j2\pi fkt}\,df
</math>
</math>
==Some Useful Fourier Transform Identities==
==Some Useful Fourier Transform Pairs==
<math>
\mathcal{F}[\alpha(t)]=\Alpha(f)
</math>
<br>
{|
|-
|<math>\mathcal{F}[c_1\alpha(t)+c_2\beta(t)]</math>
|<math>=\int_{-\infty}^{\infty} (c_1\alpha(t)+c_2\beta(t)) e^{-j2\pi ft}\, dt</math>
|-
|
|<math>=\int_{-\infty}^{\infty}c_1\alpha(t)e^{-j2\pi ft}\, dt+\int_{-\infty}^{\infty}c_2\beta(t)e^{-j2\pi ft}\, dt</math>
|-
|
|<math>=c_1\int_{-\infty}^{\infty}\alpha(t)e^{-j2\pi ft}\, dt+c_2\int_{-\infty}^{\infty}\beta(t)e^{-j2\pi ft}\, dt=c_1\Alpha(f)+c_2\Beta(f)</math>
|-
|}
<br>
<math>
\mathcal{F}[\alpha(t-\gamma)]=e^{-j2\pi f\gamma}\Alpha(f)
</math>
<br>
<math>
<math>
\mathcal{F}[\alpha(t)*\beta(t)]=\Alpha(f)\Beta(f)
\mathcal{F}[\alpha(t)*\beta(t)]=\Alpha(f)\Beta(f)
Line 96: Line 117:
\mathcal{F}[\alpha(t)\beta(t)]=\Alpha(f)*\Beta(f)
\mathcal{F}[\alpha(t)\beta(t)]=\Alpha(f)*\Beta(f)
</math>
</math>
<br>


==A Second Approach to Fourier Transforms==
==A Second Approach to Fourier Transforms==

Revision as of 11:41, 10 December 2004

From the Fourier Transform to the Inverse Fourier Transform

An initially identity that is useful: X(f)=∫−∞∞x(t)e−j2πftdt

Suppose that we have some function, say β(t), that is nonperiodic and finite in duration.
This means that β(t)=0 for some Tα<|t|

Now let's make a periodic function γ(t) by repeating β(t) with a fundamental period Tζ. Note that limTζ→∞γ(t)=β(t)
The Fourier Series representation of γ(t) is
γ(t)=∑k=−∞∞αkej2πfkt where f=1Tζ
and αk=1Tζ∫−Tζ2Tζ2γ(t)e−j2πktdt
αk can now be rewritten as αk=1Tζ∫−∞∞β(t)e−j2πktdt
From our initial identity then, we can write αk as αk=1TζB(kf)
and γ(t) becomes γ(t)=∑k=−∞∞1TζB(kf)ej2πfkt
Now remember that β(t)=limTζ→∞γ(t) and 1Tζ=f.
Which means that β(t)=limf→0γ(t)=limf→0∑k=−∞∞fB(kf)ej2πfkt
Which is just to say that β(t)=∫−∞∞B(f)ej2πfktdf

So we have that ℱ[β(t)]=B(f)=∫−∞∞β(t)e−j2πftdt
Further ℱ−1[B(f)]=β(t)=∫−∞∞B(f)ej2πfktdf

Some Useful Fourier Transform Pairs

ℱ[α(t)]=A(f)

ℱ[c1α(t)+c2β(t)] =∫−∞∞(c1α(t)+c2β(t))e−j2πftdt
=∫−∞∞c1α(t)e−j2πftdt+∫−∞∞c2β(t)e−j2πftdt
=c1∫−∞∞α(t)e−j2πftdt+c2∫−∞∞β(t)e−j2πftdt=c1A(f)+c2B(f)


ℱ[α(t−γ)]=e−j2πfγA(f)
ℱ[α(t)*β(t)]=A(f)B(f)
ℱ[α(t)β(t)]=A(f)*B(f)

A Second Approach to Fourier Transforms