| Fourier transform | |
|---|---|
| Name | Fourier transform |
| Field | Mathematics; Physics |
| Introduced | 1822 |
| Author | Joseph Fourier |
| Related | Fourier series, Laplace transform |
Fourier transform
The Fourier transform is an integral transform that decomposes functions or signals into frequencies and is central to the analysis of wave phenomena in Quantum Mechanics and Quantum Physics. In quantum theory it provides the bridge between position-space and momentum-space representations of wavefunctions, underpins the formalism of operators, and illuminates foundational results such as the Heisenberg uncertainty principle. Its mathematical structure enables both theoretical insight and practical computation across mathematical physics and experimental quantum information science.
The Fourier transform F[f](k) of an integrable function f(x) on the real line is typically defined by an integral transform named after Joseph Fourier; common conventions use factors of 2π to ensure unitarity in L²(ℝ): F[f](p) = (1/√(2πħ)) ∫_{-∞}^{∞} f(x) e^{-ipx/ħ} dx, where ħ denotes the reduced Planck constant and p is the conjugate variable. In higher dimensions, the transform generalizes over ℝ^n and is linked to the Fourier inversion theorem and the Plancherel theorem, which guarantee invertibility and preservation of inner products for square-integrable functions. Alternative formulations include the Discrete Fourier transform (DFT) and the Fast Fourier transform (FFT) algorithm used for sampled data.
In Dirac formalism, the position-space wavefunction ψ(x) and momentum-space wavefunction φ(p) form a Fourier pair: φ(p) = ⟨p|ψ⟩ = (1/√(2πħ)) ∫ ψ(x) e^{-ipx/ħ} dx. This transformation implements the unitary operator connecting the eigenbases of the position operator x̂ and the momentum operator p̂ = −iħ d/dx, reflecting the canonical commutation relation [x̂,p̂]=iħ. Foundational texts by Paul Dirac and treatments in John von Neumann's mathematical framework formalize these correspondences. The Fourier transform also relates to spectral decompositions of self-adjoint operators in the sense of the spectral theorem used in quantum Hamiltonian analysis.
The Fourier transform is linear and, with appropriate normalization, unitary on Hilbert spaces such as L²(ℝ), preserving probability amplitudes and norms; this is expressed by the Plancherel theorem and Parseval's theorem. The transform carries convolution in one domain to pointwise multiplication in the dual domain, an identity exploited in scattering theory and time evolution: (f * g)^∧ = f^∧ · g^∧. Important theorems include the Riemann–Lebesgue lemma (decay of transforms), the Paley–Wiener theorem (analytic continuation properties), and the Fourier inversion theorem ensuring reconstruction of wavefunctions from their spectral components. The transform also maps differential operators to multiplication operators, simplifying solutions of linear partial differential equations like the Schrödinger equation.
Fourier methods are ubiquitous in quantum spectral analysis: the energy spectrum of periodic systems is studied using Bloch theory and Fourier series, while scattering amplitudes and cross sections are computed via momentum-space integrals. Propagators and Green's functions for free and interacting particles are constructed using Fourier transforms of time-evolution kernels; textbooks by Richard Feynman and the formalism of Feynman path integral techniques use frequency–momentum representations. The Heisenberg uncertainty principle can be derived from basic Fourier inequalities (e.g., the Cauchy–Schwarz inequality and the Robertson–Schrödinger relation), linking spreads in position and momentum probability distributions. In many-body physics, transforms underpin the formulation of second quantization in momentum basis and are central to methods used at institutions such as CERN and Los Alamos National Laboratory.
Numerical evaluation of Fourier transforms uses discretization schemes: the Discrete Fourier transform and its efficient implementation, the Fast Fourier transform, allow computation of spectral components from sampled wavefunctions. Algorithms implemented in libraries like FFTW and in scientific computing environments (e.g., NumPy, MATLAB) support simulations of quantum dynamics, spectral density estimation, and lattice models. Careful treatment of aliasing, windowing, and sampling (via the Nyquist–Shannon sampling theorem) is required in quantum simulations on classical computers and in quantum algorithms such as the Quantum Fourier transform used in Shor's algorithm and quantum phase estimation routines implemented on platforms by IBM Quantum and Google Quantum AI.
In experimental quantum science, Fourier techniques appear in quantum state tomography and signal reconstruction from measurement data: homodyne tomography of optical fields uses inverse transforms to reconstruct Wigner functions and density matrices; this is practiced in laboratories including Max Planck Institute for Quantum Optics and university research groups. Time-of-flight measurements in cold-atom experiments infer momentum distributions via Fourier relations between initial position-space clouds and detected momentum-space images. Fourier-based filtering and deconvolution are used in quantum sensing and imaging to mitigate noise and reconstruct signals with equity-minded goals of improving access to quantum measurement capabilities in under-resourced labs. Advances in computational tomography, compressed sensing, and machine learning accelerate practical reconstruction from limited data, with ethical considerations about equitable deployment in scientific communities.