Fourier Transform

The Fourier transform expresses a function on the whole line, or on Euclidean space, as a continuous superposition of waves, exchanging the function's spatial and frequency descriptions.

Definition

The Fourier transform of a function is a new function of frequency obtained by integrating the original against complex exponentials; under suitable conditions the original is recovered by the inverse transform, making the two representations equivalent.

Scope

This topic covers the Fourier transform of integrable functions and its inversion, the interplay of smoothness and decay, the Schwartz class of rapidly decreasing functions, the Plancherel theorem on square-integrable functions, convolution and the convolution theorem, the uncertainty principle, and the extension of the transform to tempered distributions.

Core questions

• How does the Fourier transform convert between the spatial and frequency descriptions of a function?
• How are smoothness and decay of a function reflected across the transform?
• Why is the transform a unitary map on square-integrable functions?
• How does the transform turn convolution into multiplication, and why is that useful?

Key theories

Plancherel theorem

The Fourier transform extends to a unitary operator on square-integrable functions, preserving the L2 norm, so that energy is conserved between the spatial and frequency representations.

Convolution theorem and uncertainty principle

The transform turns convolution into pointwise multiplication, simplifying filtering and differential operators, while the uncertainty principle shows a function and its transform cannot both be sharply concentrated.

Clinical relevance

The Fourier transform is the central tool of signal and image processing, spectroscopy, and communications, where it analyzes frequency content and enables filtering; it diagonalizes constant-coefficient differential operators, making it indispensable for solving partial differential equations, and its discrete fast version powers modern computation.

History

The integral transform grew out of Fourier's work on heat and was placed on rigorous footing in the early twentieth century; Plancherel established its unitarity on square-integrable functions in 1910, and Schwartz's mid-century theory of distributions extended it to generalized functions.

Key figures

• Joseph Fourier
• Michel Plancherel
• Laurent Schwartz

Seminal works

• stein1971
• grafakos2008

Frequently asked questions

What is the uncertainty principle in Fourier analysis?

A function and its Fourier transform cannot both be concentrated in small regions; sharpening localization in space necessarily spreads out the frequency content, a precise inequality underlying the physical uncertainty principle.

Why does the Fourier transform help solve differential equations?

It turns differentiation into multiplication by the frequency variable, converting constant-coefficient differential equations into algebraic ones in the frequency domain that are far easier to solve.