When a wave encounters an obstacle or a medium change, it scatters. In an inverse scattering problem, we attempt to do the reverse: look at the scattered wave data and reconstruct the shape, location, and material properties of the hidden object.
However, a fundamental mathematical hurdle stands in the way: uniqueness. How do we know that two completely different objects wouldn't produce the exact same scattering pattern? If they could, the problem is fundamentally ill-posed, and true reconstruction is impossible.
Recent mathematical frameworks have revealed that uniqueness is not an elusive ideal. Instead, it can be guaranteed by exploiting specific geometric features, clever wave superpositions, and the distinct ways that physical boundaries imprint themselves onto wave fields.
Geometric Singularities as Wave Anchors
Intuitively, smooth surfaces are difficult to reconstruct because they can gently deflect waves in ways that mimic other shapes. Sharp corners, cusps, and high-curvature boundaries behave entirely differently: they act as localized singularities that force strong, non-canceling radiation patterns.
This mechanism is central to achieving uniqueness with minimal data. In Unique determination by a single far-field measurement for an inverse elastic problem, Diao et al. demonstrated that complex elastic scatterers with polygonal structures can be uniquely identified using only a single far-field measurement. By utilizing Complex Geometric Optics (CGO) to perform a corner singularity analysis, the authors showed that sharp geometric vertices prevent wave fields from being trapped or canceling out. The singularity "anchors" the wave field, ensuring its signature cannot be hidden.
Smooth Surface (Ambiguous) Polygonal Corner (Unique Signature)
~~~~~~~~~ \ /
~ ~ \ /
~ Scatters ~ \ / <-- Singularity prevents
~ smoothly ~ * wave field cancellation
~~~~~~~~~ / \
/ \
This principle explains why minimal wave data suffices for non-smooth geometries. For instance, Cheng et al., in Uniqueness in an inverse scattering problem within non-trapping polygonal obstacles with at most two incoming waves, proved that non-trapping polygonal obstacles can be uniquely identified from their far-field patterns using at most two incident waves.
When surfaces exhibit even more extreme non-standard features, such as conical singularities or cusps, standard scattering matrices fail. To resolve this, Isozaki et al. introduced a generalized scattering matrix in Conic singularities, generalized scattering matrix, and inverse scattering on asymptotically hyperbolic surfaces. This matrix accounts for geometric singularities explicitly, serving as a complete invariant that uniquely determines both the Riemannian metric and the singularity structure of the surface.
Ultimately, high curvature drives radiation. As shown by Blåsten et al. in Scattering by Curvatures, Radiationless Sources, Transmission Eigenfunctions, and Inverse Scattering Problems, high-curvature features on a scatterer's boundary inherently prevent it from becoming a "radiationless source" (an object that scatters no energy). Because high curvature forces waves to radiate, it provides a reliable geometric tool to guarantee unique shape determination.
Breaking Ambiguity in Phaseless Inverse Scattering
In practical applications like optics or X-ray crystallography, detectors can only measure the intensity (modulus) of the scattered wave, completely losing the phase information. This introduces a massive translation invariance ambiguity: shifting an object in space alters the phase of the scattered wave but leaves its intensity identical.
To achieve uniqueness without phase data, researchers use structured combinations or infinite sets of incident waves to break this symmetry.
Phaseless Problem:
Incident Wave ---> [ Object ] ---> |Scattered Wave|² (Phase Lost)
Solution (Superposition):
Wave A + Wave B ---> [ Object ] ---> Cross-terms break translation invariance
In Uniqueness in Inverse Scattering Problems with Phaseless Far-Field Data at a Fixed Frequency, Xu et al. proved that by using superpositions of multiple incident plane waves at a fixed frequency, one can overcome translation invariance. The authors linked the physical characteristics of the scattering medium to the limits of the normalized eigenvalues of the far-field operator, establishing a unique mathematical signature from intensity alone.
For infinite or open structures, such as diffraction gratings or rough terrains, an infinite sequence of waves can implicitly recover the missing phase. In Uniqueness in Inverse Diffraction Grating Problems with Infinitely Many Plane Waves at a Fixed Frequency, Hu et al. demonstrated that the shape and period of a diffraction grating are uniquely determined by phaseless near-field data. The underlying mechanism relies on analyzing the Rayleigh coefficients derived from the diffracted field expansion, which allows the full complex-valued field (including phase) to be uniquely retrieved.
Similarly, Zhang et al. proved in Uniqueness and Direct Imaging Method for Inverse Scattering by Locally Rough Surfaces with Phaseless Near-Field Data that a countably infinite set of incident plane waves over a specific measurement region provides enough constraints to uniquely lock down the shape of a locally rough surface, providing a rigorous foundation for imaging when phase measurements are impossible.
Disentangling Geometry from Substance
A final dimension of the uniqueness problem is separating where an object is from what it is. A sound-soft obstacle (which perfectly reflects waves) can produce a similar far-field pattern to a sound-hard or absorbing (Robin) obstacle of a slightly different shape. Uniqueness requires proving that geometry and material properties leave distinct, un-blendable footprints on the scattered wave.
Cheng et al. addressed this in Recovery of the shape of an obstacle and the boundary impedance from the far-field pattern by developing a two-stage framework. They proved that far-field data uniquely determines both the geometric boundary and the complex surface impedance simultaneously. They achieved this via a "probe method" that mathematically senses the boundary first, and then uses a moment method to calculate the surface properties.
This decoupling capability remains robust even when multiple obstacles with completely different boundary conditions coexist in the same medium. As shown by the authors in Recovery of multiple obstacles by probe method, a pointwise indicator function can isolate individual boundaries and determine whether each obstacle is sound-soft, sound-hard, or governed by a Robin condition without interference.
Scattered Wave Data ---> [ Probe Method ] ---> 1. Pinpoint Geometric Boundary
---> 2. Calculate Surface Impedance
This capacity to isolate material parameters from structural geometry extends to highly complex, infinite, or subsurface environments:
- Periodic Waveguides: In The factorization and monotonicity method for the defect in an open periodic waveguide, Furuya utilized a monotonicity method—comparing near-field measurements against simulated operators—to uniquely identify hidden material changes or defects within infinite repeating structures.
- Lattice Defects: In Inverse problems for locally perturbed lattices -- Discrete Hamiltonian and quantum graph, Blåsten et al. showed that collecting scattering data across the entire energy spectrum provides enough information to uniquely reconstruct both the discrete geometric structure and the physical potentials of local defects in periodic lattices.
- Internal Interfaces: In Inverse Scattering on the Line with a Transfer Condition, Currie et al. proved that reflection coefficients and discrete eigenvalues can uniquely isolate the transfer matrix of an internal "jump" or interface within a medium.
- Subsurface Fault Friction: In Stable Recovery of Coefficients in an Inverse Fault Friction Problem, de Hoop et al. extended uniqueness proofs to geophysics. They demonstrated that by modeling a subsurface fault with the Tresca friction condition, remote seismic displacement data recorded at the surface contains a mathematically complete signature to uniquely and stably reconstruct the hidden friction coefficient on the rupture plane.
By treating geometric features as wave anchors, using superpositions to bypass missing phase data, and mathematically decoupling shape from surface impedance, these frameworks demonstrate that wave scattering data holds a rigid, unambiguous mathematical mirror to physical reality.
Key Foundational Work
Recovery of the shape of an obstacle and the boundary impedance from the far-field pattern (2003) — contributed 1 concept elements to this direction.
Uniqueness in an inverse scattering problem within non-trapping polygonal obstacles with at most two incoming waves (2005) — contributed 1 concept elements to this direction.