Optical Coherent Combining in Multi-Aperture Free-Space Receivers through Atmospheric Turbulence via Optically Computed Alignment Gradients

arXiv:2610.21876 · physics.optics, physics.ao-ph · Submitted 2026-10-15 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: Physics. The day in general physics: instruments, methods and results across the field.

Kai: We're looking at the paper: "Optical Coherent Combining in Multi-Aperture Free-Space Receivers through Atmospheric Turbulence via Optically Computed Alignment Gradients".

Mira: Here's the gist of the paper. Multi-aperture reception converts spatial diversity into a coherent-combining gain, but only if the relative phase and polarization of the collected field replicas are aligned before they are summed This paper proposes a new optical-domain combining scheme in which the alignment information is computed by light itself: the ascent directions that maximize the combined output power are formed by optical interference and read out directly as differential photocurrents of balanced detection Core Concept of Optical Combining The central idea of this work is an optical-domain combiner whose per-parameter gradient of the combined output power is measured directly The starting point for this approach is a digital-domain lineage where blind gradient-ascent phase alignment (BGAPA) maximizes the combined output power with a closed-form analytical gradient The authors observe that the very same quantities are what optical interference produces in a balanced detector, so the power-ascent direction can be measured in the optical domain, before photodetection and with no digitization of the control path.

Paper summary: Kai: So, looking at "Optical Coherent Combining in Multi-Aperture Free-Space Receivers through Atmospheric Turbulence via Optically Computed Alignment Gradients," the authors are proposing a method to handle atmospheric turbulence in multi-aperture free-space receivers without needing to estimate the channel digitally.

Mira: They achieve this by using optical interference to find the optimal ascent directions for combining power, and then reading those directions out directly as differential photocurrents from balanced detection. It’s about letting the light tell you how to align before you sum everything up.

Lev: The implication for hardware is that it could simplify real-time tracking by removing the need for per-branch digital processing just to handle alignment fluctuations caused by the atmosphere.

Kai: It shifts the work from a complex estimation problem in the digital domain to a measurement problem in the optical domain, using passive field alignment as your reference point.

Mira: The paper is focused on deriving a closed-form combining gain and loss budget based on these optical measurements, which gives us concrete limits on how well this scheme performs given residual phase error and amplitude spread.

Lev: While it’s promising for reducing computational overhead, we have to remember the model’s limitations—it only handles scalar phase and real polarization rotation, so it doesn't cover every possible impairment in a complex channel.

Kai: Ultimately, this work shows a way to compute the alignment information using light itself rather than relying on external estimates or digital computation for those crucial tracking adjustments.

Conclusion: Kai: So, this paper is about using light itself to figure out how to combine signals from multiple antennas when there's atmospheric turbulence, specifically in free-space radio receivers.

Mira: It's about taking that messy atmospheric distortion and using the interference patterns between the antennas to find the best way to align them before you sum up everything.

Lev: The core idea is measuring how much power you get by tilting each antenna, and that measurement tells you exactly which way they should be pointed for maximum output.

Kai: They take a known digital method, this blind gradient ascent phase alignment thing, and show that the direction it finds is the exact same thing that optical interference creates in a balanced detector.

Mira: So instead of needing some complicated digital processing to guess the alignment, you can measure it optically right there using balanced detection and read out those power-ascent directions as simple differential photocurrents.

Lev: On real hardware, that means you don't need a whole separate processing pipeline just to keep the antennas pointed right during a storm; the alignment information comes out of the measurement itself.

Kai: The main point is that this optical domain combining lets you compute the gain and loss budget directly from these measurements, which gives us clear limits on how much noise or phase error we can tolerate.

Mira: It moves the problem from being a difficult estimation task in the digital domain to a direct measurement in the optical domain using light’s own interference.

Lev: If this works as described, it means you could have much faster, lower-overhead tracking for these multi-antenna systems because you’re not relying on iterative software adjustments to handle atmospheric phase noise.

Kai: It really shows that we can leverage the physics of optics—the way light interferes—to solve a complex signal processing problem in free space.

Mira: But it is important to remember this method only works well for scalar phase and real polarization rotation; it doesn't handle the full complexity of every kind of distortion you might see in a real channel.

Lev: For us building actual systems, the caveat there is that you still need to model those residual errors accurately because the optical measurement gives you the best possible direction, but it can’t fix everything else in the channel.

Kai: So while this is a clever way to get alignment information out of light, we still have to deal with how much error that physical measurement actually leaves behind for our final data quality.

Pengcheng Laboratory of Shenzhen 518000, China Department of Electronic and Information Engineering Harbin Institute of Technology Shenzhen 518055, China

physics.optics, physics.ao-ph

Submitted: 2026-10-15

Updated: 2026-11-05

Comments: 18 pages, 13 figures

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Importance score: 76/100

The gist: Multi-aperture reception converts spatial diversity into a coherent-combining gain, but only if the relative phase and polarization of the collected field replicas are aligned before they are summed…

Key concepts

Optical-domain combining
This is a technique where the alignment information needed for combining signals is determined using light itself, rather than relying on electronic processing. The goal is to find the optimal way to sum different received signal replicas by measuring how the total output power changes with respect to small adjustments in phase and polarization.
Gradient-ascent phase alignment (BGAPA)
This is a digital method used previously to maximize combined output power. It relies on calculating the derivative or gradient of the combined signal power. The authors found that these same gradient calculations can be performed optically using interference, allowing for real-time alignment measurement without converting signals to digital data first.
Balanced detection
This is a method used in receivers where two identical signals are mixed and then detected separately. The difference between the two resulting photocurrents reveals information about the relative phase and polarization alignment between the input signals, which is crucial for coherent combining.

Terminology

Summary

Multi-aperture reception converts spatial diversity into a coherent-combining gain, but only if the relative phase and polarization of the collected field replicas are aligned before they are summed This paper proposes a new optical-domain combining scheme in which the alignment information is computed by light itself: the ascent directions that maximize the combined output power are formed by optical interference and read out directly as differential photocurrents of balanced detection

Core Concept of Optical Combining

The central idea of this work is an optical-domain combiner whose per-parameter gradient of the combined output power is measured directly The starting point for this approach is a digital-domain lineage where blind gradient-ascent phase alignment (BGAPA) maximizes the combined output power with a closed-form analytical gradient The authors observe that the very same quantities are what optical interference produces in a balanced detector, so the power-ascent direction can be measured in the optical domain, before photodetection and with no digitization of the control path.

Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.

Improvements for AI systems

  1. Real-time gradient ascent for phase alignment can be performed directly in an optical domain using balanced detection differential currents, eliminating the need for a estimate of the channel, no deliberate perturbation, and no per-branch digital processing. This allows branches to be aligned against one shared reference—a passive tap of the combined field itself simultaneously.

  2. The system can perform closed-form combining gain and loss budget calculations over residual phase error, branch amplitude spread, and coupler excess loss by utilizing the derived formulas (5) and (6). This provides a closed-form combining-gain and loss budget over residual phase error, branch amplitude spread, and coupler-cascade excess loss.

  3. The AI system can implement fixed-gain, division-free updates for both phase and polarization loops using the calibrated loop gains Kphi and Ktheta derived from the physical components. This ensures no dynamic dividers, no regularization floors against division by zero, and no high-speed sampling during runtime operation.

Sources

Related papers