Quantum Harmonic Analysis

• Quantum harmonic analysis is an operator-theoretic extension of classical harmonic analysis that replaces functions with operators and convolution with quantum convolution.
• It unifies time-frequency analysis, operator theory, and noncommutative geometry through the use of phase-space translations and the Fourier–Wigner transform.
• Applications span mathematical physics, data analysis, and quantum information by enabling refined spectral synthesis and localization methods.

Quantum harmonic analysis (QHA) is an operator-theoretic extension of classical harmonic analysis in which functions are systematically replaced by operators and classical convolution by noncommutative (quantum) convolution structures on suitable operator algebras. Fundamental to QHA is the use of phase-space translations and modulations for operators, mapping between function spaces and operator spaces via projective unitary representations of (often abelian) groups. The analytic machinery recovers, unifies, and extends many tools in time-frequency analysis, operator theory, noncommutative geometry, and mathematical physics, offering deep structural theorems for operators and their interaction with classical harmonic-analytic frameworks.

1. Foundations of Quantum Harmonic Analysis

At the core of QHA is the translation of classical harmonic concepts into the operator-theoretic setting. On a Hilbert space $H$ (typically $L^2(\mathbb R^d)$), the basic translation for a trace-class operator $S$ is given by time–frequency (Weyl–Heisenberg) shifts $\alpha_z(S) = \pi(z) S \pi(z)^*$, with $\pi(x,\omega)g(t) = e^{2\pi i t\cdot\omega} g(t-x)$. The convolution of two trace-class operators is the function $$(S \star T)(z) = \mathrm{tr}[S\,\alpha_z(\check{T})]$$, where $\check{T} = P T P$ and $P$ is the parity operator.

The canonical quantum Fourier transform is the Fourier–Wigner transform: $$\mathcal F_W(S)(z) = e^{-\pi i x\cdot\omega} \mathrm{tr}[\pi(-z)\,S],$$ which satisfies

$$\mathcal F_W(S\star T) = \mathcal F_W(S)\,\mathcal F_W(T).$$

This operator convolution generalizes the classical convolution-product duality under the Fourier transform, and establishes QHA's commutative Banach algebra structure encompassing both $L^1$-functions and trace-class operators (Fulsche et al., 2023).

A further crucial ingredient is the short-time Fourier transform (STFT) $V_gf(z) = \langle f, \pi(z)g\rangle$, whose integrability defines the modulation spaces $M^p(\mathbb R^d)$. For example, $M^1(\mathbb R^d)$ is a Banach algebra under pointwise multiplication and convolution, and all its members are continuous and vanish at infinity (Doerfler et al., 23 Sep 2025).

2. Quantum Convolutions, Translations, and Operator Spaces

QHA distinguishes multiple forms of convolution and translation, expanding the interplay between functions and operators:

• Function–operator convolution: For $f\in L^1(\mathbb R^{2d})$ and $A$ trace-class,

$f \star A = \int f(z)\,\alpha_z(A)\,dz.$

• Operator–operator convolution: For $S,T$ trace-class,

$$(S \star T)(z) = \mathrm{tr}[S\,\alpha_z(\check{T})].$$

• Noncommutative convolution: On general locally compact groups $G$, given a square-integrable projective unitary representation $\sigma$, the relevant convolution structures (with dual Young’s inequalities) extend to Schatten–class operators and $L^p$-spaces (Halvdansson, 2022, Sababe et al., 8 Apr 2025).

The Fourier–Wigner transform and symplectic Fourier transform intertwine operator convolution and pointwise products, producing a precise analog of Gelfand theory for operator algebras. For instance, the Banach algebra $L^1(\mathbb R^{2n}) \oplus \mathcal T^1$, equipped with QHA convolution, has maximal ideal space $\mathbb R^{2n} \times \{0,1\}$, reflecting classical/quantum duality (Berge et al., 2023).

QHA also extends naturally to lattices, in which discrete convolution structures underpin the theory of Gabor multipliers (via sampled operator translations and Fourier series), and to general abelian phase spaces with Heisenberg multipliers (Skrettingland, 2019, Fulsche et al., 2023).

3. Operator Modulation Spaces and Time-Frequency Analysis

QHA supports a rich scale of operator-valued modulation spaces. Operator modulation spaces $\mathcal M^{p,q}$ for operators $T$ are defined by the integrability of their operator-valued short-time Fourier transforms $Q_S T(w,z)$ under a projective representation of double phase-space: $\gamma_{w,z}(T) = \pi(z) T \pi(w)^*,$

$$Q_S T(w,z) = \langle T, \gamma_{w,z}(S)\rangle_{HS},$$

with $S$ chosen as a window. A central result is that modulation spaces of operators (e.g., Feichtinger's operator class $\mathcal M^{1,1}$) coincide with operators whose Weyl symbols lie in $M^1$ (Luef et al., 2024).

Discrete operator Gabor frames, formed by translates and modulates $\gamma_{(\lambda,\mu)}(S)$ over a lattice, yield atomic decompositions for operator spaces, generalizing the classical Gabor transform to operators (Luef et al., 2024). This viewpoint unifies classical pseudodifferential operator theory, Cohen’s class representations, and time-frequency localization in a single operator-theoretic context.

4. Quantum Harmonic Analysis on Groups and Homogeneous Spaces

QHA techniques generalize powerfully to non-Euclidean settings, particularly locally compact abelian groups and homogeneous spaces. For a locally compact abelian group $G$ with dual $\widehat G$, QHA uses projective unitary representations with associated multipliers $m(x, y)$ (implementing the Heisenberg condition for Pontryagin duality), and the quantum Fourier transform

$\mathcal F_U(T)(\xi) = \mathrm{tr}[T U_\xi^*],$

with inversion and spectral decompositions analogous to the Euclidean case (Fulsche et al., 2023, Mensah, 18 Sep 2025). In these settings, one still obtains Plancherel, Riemann–Lebesgue, and Wiener–Tauberian theorems for operator convolutions and for Banach algebras including both function and operator components.

When $G$ is non-unimodular, a Duflo–Moore operator is introduced to correct for the modular function, and admissibility conditions for operators are characterized by integrability under group translation. This generalization encompasses the standard Weyl–Heisenberg group, affine groups, and further non-abelian contexts (Halvdansson, 2022, Berge et al., 2021).

For function spaces such as the Bergman space over the unit ball or polyanalytic Fock spaces, QHA provides descriptions of Toeplitz algebras, explicit Berezin transforms, and Tauberian theorems, clarifying the operator-theoretic underpinnings of complex analysis on symmetric domains (Dawson et al., 2024, Fulsche et al., 2023).

5. Spectral Synthesis, Segal Algebras, and Structural Theorems

QHA supports a full theory of quantum Segal algebras—dense Banach subalgebras of the combined function/operator ambient algebra, characterized by shift-invariance and Banach algebra properties. Prominent subclasses include Feichtinger’s algebra of operators (operators whose Weyl symbol lies in the classical Feichtinger algebra) and operator quantizations of classical Segal algebras (Berge et al., 2023, Sababe et al., 8 Apr 2025).

The spectral theory mirrors classical harmonic analysis: zeros in the operator Fourier transform control regularity, density, and Tauberian phenomena. Spectral synthesis is established for quantum Segal algebras, guaranteeing the recoverability of closed ideals from their Fourier zeros, and quantum Wiener approximation theorems provide polynomial density in the combined algebra (Sababe et al., 8 Apr 2025).

Operator-theoretic Gelfand theory is robust, with the spectrum reflecting symplectic or group-theoretic dualities. For instance, the Gelfand maximal ideal space for $L^1 \oplus \mathcal T^1$ is $\mathbb R^{2n} \times \{0,1\}$.

6. Applications in Data Analysis, Noncommutative Geometry, and Mathematical Physics

Contemporary research applies QHA to manifold learning, data augmentation, and feature extraction. A principal result is that augmentation via operator-convolution produces principal components in $M^1(\mathbb R^d)$: such eigenfunctions are continuous, rapidly decaying, and well-adapted for stable kernel constructions in manifold learning algorithms. Empirical studies demonstrate enhanced time-frequency smoothness and localization for principal components of augmented datasets compared to their unaugmented counterparts (Doerfler et al., 23 Sep 2025).

In the field of operator PDEs, QHA supports functional spaces such as spectral Barron spaces, complete with continuous embeddings, interpolation properties, and contraction-mapping solutions for operator equations (e.g., Schrödinger-type equations) (Mensah, 18 Sep 2025).

Noncommutative geometry is deeply interwoven with QHA. Quantum Segal algebras serve as natural involutive algebras for spectral triples with Dirac-type operators, enabling index theorems in the sense of Connes. QHA functional calculi mediate between noncommutative geometry (cyclic cohomology, Fredholm modules) and quantum harmonic representations for invariants such as the quantized Hall conductance (Sababe et al., 8 Apr 2025).

In quantum information theory, QHA constructs families of quantum channels by representing measure and Fourier multiplier algebras on group-invariant operator spaces, providing explicit counterexamples to the asymptotic quantum Birkhoff conjecture, and linking channel capacities, error correction, and entanglement-breaking properties to harmonic-analytic data (Crann et al., 2012).

7. Broader Impact and Open Directions

QHA thus constitutes a comprehensive analytical framework, integrating classical and noncommutative harmonic analysis, time-frequency analysis, operator algebras, noncommutative geometry, and mathematical physics. Its techniques rigorously ground pseudodifferential and Toeplitz operator theory, operator-valued wavelet transforms, phase-space localization, and modern forms of spectral approximation.

Open research directions span:

• Extending QHA to fully non-abelian groups, quantum groups, and other noncommutative geometries.
• Classifying interpolation and embedding properties for QHA-based $L^p$-theories.
• Developing quantum uncertainty, Sobolev, and Lieb–Thirring estimates for operator phase-space functions.
• Advancing quantum Besov, Triebel–Lizorkin, and coorbit space frameworks.
• Expanding applications to quantum signal processing, statistical learning, and topological phases of matter (Sababe et al., 8 Apr 2025, Doerfler et al., 23 Sep 2025).

QHA forms a foundational bridge between algebraic, analytic, and geometric approaches to operators, phase space, and quantum systems, with ongoing influence across mathematical analysis, physics, and data science.