Scribd
Upload
30 views

1 The Fourier Transform

The document discusses the Discrete Fourier Transform (DFT), which is a discrete version of the continuous Fourier Transform that maps a discrete set of numbers to another discrete set of numbers. The DFT is more efficient for computation than the continuous Fourier Transform. It can be represented by a matrix that is symmetrical and has orthogonal columns. The Fast Fourier Transform (FFT) further improves the efficiency of the DFT to O(n log n) by recursively factorizing the DFT matrix.

Uploaded by

proust0
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
30 views

1 The Fourier Transform

The document discusses the Discrete Fourier Transform (DFT), which is a discrete version of the continuous Fourier Transform that maps a discrete set of numbers to another discrete set of numbers. The DFT is more efficient for computation than the continuous Fourier Transform. It can be represented by a matrix that is symmetrical and has orthogonal columns. The Fast Fourier Transform (FFT) further improves the efficiency of the DFT to O(n log n) by recursively factorizing the DFT matrix.

Uploaded by

proust0
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 3

The Discrete Fourier Transform

Didi Park
13 May 2014
1 The Fourier Transform
Let us dene the Fourier transform

f, f : R C to be

f() =

f(x)e
2ix
dx, for real .
f(x) must be square integrable, that is,

f(x)
2
dx < .
The Fourier transform also has an easy inverse:
f(x) =

f()e
2ix
d.
Speaking in linear algebra terms, the Fourier transform is a linear transformation.
This can easily be seen from the linearity and scalar multiplicity of the integral. Both
the domain and codomain, f and

f, eectively form innite-dimensional vector spaces.
1.1 Applications
The Fourier transform relates a functions time or spatial domain to a frequency domain.
It can be used to solve a wide variety of problems, including the following:
Partial (and ordinary) dierential equations
Quantum mechanics
Signal processing
Data compression
1
2 The Discrete Fourier Transform
As you can see, the Fourier transform is continuous, making it dicult for computers
to perform the transform to high accuracy. It also requires that the function being
transformed be square integrable, thus limiting the possible inputs signicantly.
The discrete fourier transform (DFT) is discrete, making it much more conducive to
computation. Unlike the Fourier transform, it maps a countable set of numbers to a new
countable set of numbers. If youre only interested in transforming a small set of data,
use the DFT. Another consequence of its discrete nature is that it can be used to map
points from any function, square integrable or not.
2.1 Denition
We simply discretize the Fourier transform by changing the integral to a sum, appropri-
ately:
X
k
=
N1

n=0
x
n
e
2in
k
N
.
Note that the
k
N
serves the same purpose as in the continuous case.
2.2 Matrix representation
The DFT, unsurprisingly, can be written as a matrix transformation.

1 1 1 . . . 1
1 w w
2
. . . w
N1
1 w
2
w
4
. . . w
2(N1)
.
.
.
.
.
.
.
.
. . . .
.
.
.
1 w
N1
w
2(N1)
. . . w
(N1)
2

where w = e
i
2
N
.
Notice that this matrix is symmetrical. In fact, its columns are also orthogonal (keep in
mind that the complex dot product requires one to take the conjugate transpose).
3 The Fast Fourier Transform
The DFT has eciency O(n
2
). The fast Fourier transform (FFT) takes advantage of
the symmetry and orthogonality of the DFT matrix, and recursively factors it until we
reach an eciency of O(nlog(n)). Let us take the example of the 8 8 DFT matrix, F
8
.
F
8
=

1 D
1 D

2
4 Further reading/Works Cited
1. Strang, Gilbert. Lec 26, MIT 18.06 Linear Algebra. https://www.youtube.com/
watch?v=M0Sa8fLOajA
2. Tao, Terrence. An Epsilon of Room, I. https://terrytao.files.wordpress.
com/2012/12/gsm-117-tao3-epsilon1.pdf.
3. Terras, Audrey. (1999). Fourier Analysis on Finite Groups and Applications.
4. Wikipedia. DFT Matrix. https://en.wikipedia.org/wiki/DFT_matrix.
5. Wikipedia. Fourier Transform. https://en.wikipedia.org/wiki/Fourier_transform
3

You might also like