Gabor Wave Front Set Analysis
- The Gabor wave front set is a global phase-space tool that localizes and characterizes singularities in distributions through STFT decay properties.
- It employs anisotropic scaling to capture direction-dependent phenomena in dispersive and non-isotropic evolution problems.
- Its invariance properties, including metaplectic invariance and equivalence to classical wave front sets, ensure robust microlocal analysis in both continuous and discrete settings.
The Gabor wave front set is a global, phase-space microlocal object that characterizes the localization and singularities of generalized functions or distributions, using time-frequency analysis via the short-time Fourier transform (STFT) and Gabor frames. It refines the classical microlocal analysis by encoding precise joint space-frequency information about singularities, and its anisotropic variants introduce further flexibility to track directionally dependent phenomena such as those arising in dispersive equations and evolution problems with non-isotropic scaling.
1. Time-Frequency Foundations and Core Definitions
Let be a tempered distribution. The central analytic tool is the short-time Fourier transform (STFT), defined via a nonzero Schwartz window as
where is modulation by frequency , is translation by . The STFT quantifies the local behavior of near position and frequency , exhibiting rapid decay if is smooth.
A point is not in the Gabor wave front set if there exists an open conic neighborhood such that, for every ,
with . Equivalently, rapid decay of the STFT on cones in phase space signals microlocal regularity, and records the failure of this property.
This definition is independent of the choice of window and is equivalent in both the continuous and discrete (Gabor frame) settings (Rodino et al., 2012, Rodino et al., 2020).
The anisotropic Gabor wave front set for adapts the notion by introducing space-frequency anisotropy: for anisotropic dilation , (Wahlberg, 2023).
2. Structural Properties and Equivalence to Classical Notions
The Gabor wave front set exhibits several key properties:
- Closedness and conic structure: is a closed conic subset of (Rodino et al., 2012, Rodino et al., 2020).
- Metaplectic invariance: Symplectic (linear canonical) transforms lift to metaplectic operators on ; for (Rodino et al., 2020).
- Microlocality: For Weyl quantizations with in an appropriate global symbol class (e.g., Shubin ), and does not generate new singularities outside (Rodino et al., 2012, Rodino et al., 2020).
- Characterization of smoothness: (Rodino et al., 2012).
Fundamentally, coincides with Hörmander’s global wave front set defined via Shubin-class pseudodifferential localization. The precise equality for is due to Rodino and Wahlberg (Rodino et al., 2012, Rodino et al., 2020, Schulz et al., 2013). In the anisotropic setting, reduces to the classical case at .
The homogeneous wave front set introduced by Nakamura, which uses semiclassical scaling, is also shown to be equivalent to for tempered distributions (Schulz et al., 2013).
3. Anisotropic Gabor Wave Front Set
The anisotropic Gabor wave front set , and more generally for a rational anisotropy parameter , adapts microlocal analysis to non-isotropic scalings. This is necessary for problems where physical or geometric anisotropy is present, such as in higher-order Schrödinger or dispersive equations with polynomial principal symbols.
The definition assigns different scaling behaviors to and : , and the STFT is probed along such dilations. The resulting set is conic with respect to this -anisotropic scaling and retains the invariance and closedness properties of the isotropic case (Cappiello et al., 2023, Wahlberg, 2023).
4. Propagation of Gabor Singularities
A central application is the propagation of singularities under evolution equations, particularly those of Schrödinger or dispersive type. For a broad class of linear equations with real-valued polynomial symbols or Weyl quantizations belonging to anisotropic symbol classes , the propagation of the (anisotropic) Gabor wave front set aligns with the characteristic Hamiltonian flow:
where is the Hamiltonian flow generated by the principal symbol, respecting the anisotropic scaling (Cappiello et al., 2023, Wahlberg, 2023).
For operators whose Schwartz kernels satisfy a graph-type conic support condition (see below), the action on singularities is governed by a precise microlocal relation: with the relation (Wahlberg, 2023).
Examples:
- For the free Schrödinger equation (, ), .
- For fourth-order evolution (, ), singularities propagate along (Wahlberg, 2023).
5. Graph-Type Criterion for Operator Kernels
A distinguishing feature of the Gabor microlocal framework is that propagation laws extend to operators with non-smooth or generalized kernel distributions. The key sufficient hypothesis is a graph-type conic support criterion for the anisotropic Gabor wave front set of the Schwartz kernel :
Define two sets associated with : If both are empty, then must be contained microlocally near the graph of an invertible linear map , ensuring that is well-behaved with respect to phase space localization and allows continuous extension to (Wahlberg, 2023).
This abstraction supports propagation and restriction theorems for evolution equations and their associated propagators.
6. Examples and Explicit Computations
Several canonical distributions illustrate the theoretical structure (Rodino et al., 2020, Rodino et al., 2012, Boiti et al., 2017):
| Distribution | STFT Behavior | |
|---|---|---|
| Dirac | Non-decay in frequency | |
| Constant $1$ | Non-decay in space | |
| Plane wave | Localized along frequency | |
| Chirp | Non-decay along |
For compactly supported distributions , the Gabor wave front set is , i.e., all singularities are concentrated at in phase space with the same frequency projection as the classical wave front set. Conversely, functions have empty (Wahlberg, 2019).
7. Extensions and Generalizations
Multiple frameworks broaden the scope of the Gabor wave front set:
- Modulation spaces and ultradifferentiable classes: In the presence of a non-quasianalytic weight , the Gabor -wave front set is defined by exponential (rather than polynomial) decay of weighted by . Equivalence with decay-based definitions and Gabor frame samples holds, providing robust ultradifferentiable microlocality (Boiti et al., 2017).
- Stability under frame perturbations: The wave front set determined via a Gabor frame is invariant under -type perturbations and under nonstationary Gabor frames, confirming the robustness of under reasonable changes to the time-frequency lattice or window (Boiti et al., 16 Jan 2026).
- Analytic, Gevrey, and modulation space variants: Defining wave front sets by requiring decay of at an exponential or modulation-space rate allows fine-grained spectral and analytic microanalysis (Rodino et al., 2020).
These generalizations have facilitated micro-analysis for a broad spectrum of differential and pseudo-differential operators, including those with limited regularity, complex symbols, or global (nonlocal) effects.
The Gabor wave front set and its anisotropic extensions unify microlocal and time-frequency analysis, providing a geometrically natural, robust, and computationally accessible microlocal structure for distributions. It underpins exact propagation laws for singularities and enables precise control in ultra- and analytic function settings (Rodino et al., 2020, Rodino et al., 2012, Wahlberg, 2023, Cappiello et al., 2023, Boiti et al., 2017, Boiti et al., 16 Jan 2026).