25.08.2021 Views

082-Engineering-Mathematics-Anthony-Croft-Robert-Davison-Martin-Hargreaves-James-Flint-Edisi-5-2017

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

788 Chapter 24 The Fourier transform

inthe following way:

∞∑

˜f(t)=T f[n]δ(t −nT)

n=0

Notethatwhereas f (t)istheoriginalcontinuousfunctionoft, ˜f (t)isanapproximation

obtainedusingonlythesampledvalues.Thisrepresentationisdiscussedingreaterdetail

inAppendix I.

24.10.2 Derivation

Thisderivationisbaseduponthefactthatafunction f (t),definedfort 0,andhaving

been sampled atintervals ofT, can be represented by the function ˜f (t) where

∞∑

˜f(t)=T f[n]δ(t −nT)

n=0

In any real problem it is only possible to sample a signal over a finite time interval.

Suppose we obtainN samples of the signal at timest = 0,T,2T,..., (N −1)T. Then

thesampledsignalcanberepresentedbyamendingthelimitsofsummationandwriting

˜f(t)=T

∑N−1

n=0

f[n]δ(t −nT)

Taking the Fourier transform of both sides gives

F{ ˜f(t)}=T

∫ ∞

−∞

∑N−1

e −jωt

n=0

f[n]δ(t −nT)dt

If we make the assumption that it is permissible to interchange the order of integration

and summation wefind

F{ ˜f(t)}=T

=T

∑N−1

f[n]

∫ ∞

n=0

−∞

∑N−1

f[n]e −jωnT

n=0

e −jωt δ(t −nT )dt

where we have used the sifting property of the delta function. The quantity on the r.h.s.

isacontinuousfunctionof ωderivedusingthevaluesinthesequence f[n].Writethisas

F(ω)andnotethat ˜ F(ω)isanapproximationtotheFouriertransformF(ω) ˜

= F{f (t)}.

It can also be thought of as a Fourier transform for sequences. An important point to

note about this function of ω is that it is periodic with period 2π . This is proved in the

T

following example.

Example24.21 Show thatthe function

∑N−1

F(ω) ˜ =T

n=0

f[n]e −jωnT

isperiodic with period 2π

T .

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!