Fourier Transform — Mathematics Knowledge System

Overview

The Fourier transform extends Fourier series from periodic functions to arbitrary integrable functions on the real line. Where a Fourier series decomposes a periodic function into discrete harmonics, the Fourier transform decomposes a general function into a continuous spectrum of frequencies. It is one of the most powerful tools in analysis, with applications spanning signal processing, PDEs, quantum mechanics, and number theory.

Intuition

If a Fourier series is a recipe for building a periodic sound from pure tones, the Fourier transform is the recipe for a one-time sound -- an explosion, a spoken word, a heartbeat pulse. The transform tells you the amplitude and phase of every frequency present in the signal, giving a complete 'frequency fingerprint'. Taking the inverse transform reconstructs the original signal perfectly.

Formal Definition

Definition

For f in L^1(R), the Fourier transform is the function of frequency xi given by:

\hat{f}(\xi) = \int_{-\infty}^{\infty} f(x)\, e^{-2\pi i x \xi}\, dx

Fourier transform (frequency convention)

f(x) = \int_{-\infty}^{\infty} \hat{f}(\xi)\, e^{2\pi i x \xi}\, d\xi

Inverse Fourier transform

\hat{f}(\xi) = \mathcal{F}[f](\xi), \quad f(x) = \mathcal{F}^{-1}[\hat{f}](x)

Operator notation

Notation

Notation	Meaning
$\hat{f}(\xi)$	Fourier transform of f at frequency xi
$\mathcal{F}$	Fourier transform operator
$\xi$	Frequency variable (cycles per unit length)
$\omega = 2\pi\xi$	Angular frequency (radians per unit length)

Theorems

Theorem 1: Plancherel Theorem

\|\hat{f}\|_{L^2} = \|f\|_{L^2}, \text{ i.e., } \int |\hat{f}(\xi)|^2\,d\xi = \int |f(x)|^2\,dx

Theorem 2: Convolution Theorem

\widehat{f * g}(\xi) = \hat{f}(\xi)\,\hat{g}(\xi), \text{ where } (f*g)(x) = \int f(y)g(x-y)\,dy

Theorem 3: Fourier Inversion

\text{If } f, \hat{f} \in L^1(\mathbb{R}), \text{ then } f(x) = \int \hat{f}(\xi)\,e^{2\pi i x\xi}\,d\xi \text{ a.e.}

Theorem 4: Derivative Rule

\widehat{f'}(\xi) = 2\pi i \xi\, \hat{f}(\xi)

Worked Examples

1
Complete the square in the exponent after substituting u = x + i*xi:
$\hat{f}(\xi) = \int_{-\infty}^{\infty} e^{-\pi x^2} e^{-2\pi i x\xi}\,dx = e^{-\pi\xi^2}\int_{-\infty}^{\infty} e^{-\pi(x+i\xi)^2}\,dx$
2
The Gaussian integral equals 1 by contour shift, giving:
$\hat{f}(\xi) = e^{-\pi\xi^2}$

✓ Answer

The Gaussian is its own Fourier transform: the transform of e^{-pi x^2} is e^{-pi xi^2}.

Practice Problems

Mediumfree response

State Parseval's theorem (Plancherel) and explain its physical interpretation for a signal f(t).

Mediumproof writing

Use the derivative rule to solve the ODE f'(x) + 2pi*f(x) = delta(x) using the Fourier transform.

Common Mistakes

Common Mistake

Thinking the Fourier transform and Fourier series are the same

Fourier series are for periodic functions (discrete spectrum). The Fourier transform is for general L^1 or L^2 functions (continuous spectrum). The series is a special case in the distributional sense.

Common Mistake

Confusing the various sign/factor conventions

There are three common conventions differing in placement of the 2pi factor: in the exponent, in front of the integral, or split as 1/sqrt(2pi) in both. The choice affects the convolution theorem and Plancherel constant -- always state which convention you use.

Quiz

Historical Background

Building on Fourier's series work (1822), the Fourier transform emerged from the need to analyze non-periodic signals. Cauchy and Poisson used integral transforms in the early 19th century. The rigorous L^2 theory was developed by Plancherel (1910), who proved the transform is an isometry on L^2(R). Shannon's sampling theorem (1949) and Cooley-Tukey's Fast Fourier Transform (1965) made the transform central to digital signal processing.

1822
Fourier publishes Theorie analytique de la chaleur
Joseph Fourier
1910
Plancherel proves the L^2 isometry (Plancherel theorem)
Michel Plancherel
1949
Shannon's sampling theorem published
Claude Shannon
1965
Cooley-Tukey Fast Fourier Transform algorithm published
James Cooley, John Tukey

Summary

The Fourier transform decomposes a function into a continuous spectrum of frequencies: f-hat(xi) = integral f(x) e^{-2pi i x xi} dx.
The inverse transform reconstructs f from its frequency content; inversion holds whenever f and f-hat are both in L^1.
Plancherel theorem: the transform extends to an isometry of L^2(R), preserving the L^2 norm.
Convolution theorem: F[f * g] = F[f] F[g] -- convolution in time equals multiplication in frequency.
Derivative rule: F[f'](xi) = 2pi i xi F[f](xi) -- differentiation becomes multiplication, making the FT a key tool for PDEs.

References

BookStein, E. and Shakarchi, R. -- Fourier Analysis: An Introduction (2003), Princeton University Press
BookRudin, W. -- Real and Complex Analysis (3rd ed., 1987), Chapter 9
WebsiteWikipedia -- Fourier transform