In classical optics and wave physics, scattering is traditionally viewed as an enemy. When light travels through a turbulent atmosphere, a foggy window, or a complex biological tissue, multiple scattering scrambles the wavefronts, turning coherent signals into seemingly random noise and severely degrading image quality.
To circumvent this, conventional imaging systems rely on clear, homogeneous media where waves travel along predictable, straight paths. In these settings, the spatial resolution of an imaging system is strictly bound by the classical diffraction limit: it is governed by the wavelength of the wave and the physical size of the collection lens, or aperture. Waves propagating at highly oblique angles miss the detector entirely, carrying away high-frequency spatial information forever.
However, a profound insight developed across acoustics, seismology, and optics turns this paradigm on its head. By leveraging techniques like time reversal (TR), researchers discovered that a highly scattering, random medium does not destroy information—it redistributes it. Instead of restricting resolution, multiple scattering can act as a massive, complex "virtual lens" that effectively expands the physical aperture of a detector, enabling wave focusing and imaging that shatters the classical diffraction limit.
The Mechanism: How Scattering Builds a Virtual Lens
To understand why this works, consider a standard time-reversal experiment. A localized source emits a short wave pulse. In a homogeneous medium, the wave expands outward. A small array of transceivers—a time-reversal mirror (TRM)—captures a small slice of this wavefront, records it, reverses the signal in time, and plays it back. Because the TRM only captures a narrow aperture of the total wave, the refocused spot at the original source location is relatively wide, limited by the geometry of the array.
Now, place a complex, randomly scattering medium between the source and the TRM. As the wave travels forward, it undergoes countless reflections and refractions, a phenomenon known as multipathing. Waves that would have completely bypassed the physical boundaries of the TRM are scattered back into it.
As demonstrated by Kent D. Taylor et al. in Statistical stability in time reversal, this multipathing effect can be modeled mathematically as an effective expansion of the sensor array's physical dimensions. The random environment acts as a virtual aperture. When the scrambled, prolonged signal is time-reversed and transmitted back through the same medium, the waves retrace their intricate paths in reverse. The scattering events that originally dispersed the energy now perfectly undo the distortion, funneling the wave components from wide, oblique angles back to the source.
Homogeneous Medium:
Source ⊙ ─────────────────> [ Small Physical Aperture ]
(Oblique waves escape into space)
Random/Scattering Medium:
Source ⊙ ─── \ ─── / ─────> [ Small Physical Aperture ]
\ /
\ /
(Virtual Aperture captures oblique paths via scattering)
This mechanism yields a significantly tighter focal spot than what is possible in a clean environment. In Time-Reversal Aperture Enhancement, Jean‐Pierre Fouque et al. proved that random fluctuations compensate for the limitations of a small physical mirror. They quantified this super-resolution effect by showing that the diffracted field generated by time reversal in randomly layered media decays at a rate of $1/|x|^4$, compared to the much slower $1/|x|$ decay observed in homogeneous media, indicating a highly localized and precise focus.
The Propagation-Resolution Duality
In a clear medium, moving a detector further away from a source degrades spatial resolution because the collection angle (numerical aperture) shrinks. In a scattering medium, an entirely opposite, counter-intuitive scaling law emerges.
Because resolution depends on the diversity of the paths captured by the virtual aperture, more scattering events lead to a larger virtual lens. In Superresolution and duality for time-reversal of waves in random media, Albert Fannjiang et al. established a formal mathematical duality between the forward spatial spreading of a wave and its resolution upon time reversal. They formulated an uncertainty inequality demonstrating that increased wave diffusion in a complex medium directly acts as a resource for higher-resolution refocusing.
This duality leads to two remarkable properties:
- Distance-Enhanced Resolution: As a wave propagates further through a scattering or turbulent medium, it encounters more random structures, expanding its angular spread. Consequently, the resolution of the refocused spot can actually improve with increased propagation distance, as noted by Kent D. Taylor et al. and further analyzed in the context of atmospheric turbulence by Albert Fannjiang et al.
- Aperture Independence: When the scattering properties of the medium dominate the wave dynamics, the ultimate resolution of the refocused wave becomes entirely independent of the physical aperture size of the recording device. The medium itself dictates the information capacity of the wave field.
Why Is It Stable?
A natural skepticism arises: if this super-resolution relies on the exact, chaotic microscopic configuration of a random medium, wouldn't the slightest change or measurement error destroy the effect?
The reason this mechanism works reliably in practice is a property known as statistical stability (or self-averaging). In Statistical stability in time reversal, the authors proved that provided the source is spatially localized or the signal possesses sufficient frequency bandwidth, the refocused focal spot is statistically stable. This means that the super-resolution effect does not depend on the specific, microscopic realization of the random fluctuations. The massive number of independent scattering paths averages out the fluctuations, ensuring a robust, reproducible focal spot.
Generalizing Beyond Time-Reversal Mirrors
While initially studied in acoustic time-reversal mirrors, the principle that scattering media redistribute and preserve high-frequency spatial information has spread across multiple domains:
- Random Waveguides: In Paraxial Coupling of Propagating Modes in Three-Dimensional Waveguides with Random Boundaries, Liliana Borcea demonstrated that stochastic boundary fluctuations (such as rough walls or ocean surfaces) couple propagating modes together, creating a virtual aperture that beats standard waveguide imaging limits. Later, Liliana Borcea showed in A ghost imaging modality in a random waveguide that intense scattering distributes target information across all available propagation modes, making image reconstruction highly robust and less dependent on physical detector sizes.
- Sub-Diffraction Imaging and Microscopy: The concept of using local scattering structures to capture sub-diffraction information is utilized in optical microscopy. For instance, Haibo Yu et al. in Scanning superlens microscopy for non-invasive large field-of-view visible light nanoscale imaging utilized high-refractive-index microspheres to convert near-field evanescent waves—which normally decay rapidly and vanish—into propagating far-field waves, achieving nanoscale resolution with visible light.
- Computational Decoding: Modern computational imaging seeks to explicitly decode these scrambled signals. As highlighted by Shwetadwip Chowdhury et al. in Supplementary document for High-resolution 3D refractive index microscopy of multiple-scattering samples from intensity images - 4044047.pdf, multiple scattering is increasingly treated not as an obstacle, but as a computational opportunity to extract high-frequency structural information that a clear medium would fail to transmit.
- Radar and Geophysics: Similar principles apply to radar imaging. In Imaging of the Internal Structure of Permafrost in the Tibetan Plateau Using Ground Penetrating Radar, Yao Wang et al. demonstrated that Reverse Time Migration (RTM) algorithms can actively exploit diffracted and scattered energy to accurately map complex, underground permafrost structures that would otherwise be obscured under conventional imaging frameworks.
By shifting our perspective from treating scattering as a source of destructive noise to recognizing it as a complex, high-capacity physical lens, this framework provides a powerful methodology for achieving super-resolution across diverse wave-based imaging modalities.
Key Foundational Work
Time-Reversal Aperture Enhancement (2003) — contributed 4 concept elements to this direction.
Statistical stability in time reversal (2002) — contributed 3 concept elements to this direction.