Key Takeaways
- Quantum communications face significant challenges due to the mixed properties of real-world channels, making it difficult for systems to process information accurately.
- Researchers at INRS developed a method using the Talbot Array Illuminator, improving signal quality nearly tenfold by denoising quantum signals in noisy environments.
- This new technique utilizes standard telecom components, making it possible to integrate quantum communications with existing infrastructure.
- The method effectively redistributes quantum correlations into tight peaks while discarding incoherent noise, allowing for clearer quantum state recovery.
- Future applications could extend to satellite communication, requiring further testing but holding promise for practical quantum networking solutions.
The photon is there, somewhere. Paired with its partner, encoded with quantum information, travelling through a fibre-optic cable or bouncing off a satellite mirror toward a detector on the ground. The problem is that it is travelling in the company of millions of other photons, most of them noise, stray light from the sun or from amplifiers in the network, and to the detector they all look roughly the same. The quantum state, which is in a very real sense the whole point, is buried. Unreadable. In practical terms, it’s lost.
This is perhaps the central unsolved problem in quantum communications right now, and it is what makes most demonstrations of quantum networks so fussy, so fragile, so thoroughly confined to the lab. Real-world channels are bright, messy places, quite unlike the temperature-controlled dark-fibre environments where entangled photons currently do most of their work.
A team at the Institut national de la recherche scientifique in Montreal has now come up with a surprisingly compact solution. Their approach, published in Science Advances, uses a piece of classical optical kit called a Talbot Array Illuminator, repurposed in a way that nobody had really thought to try before: applied not to beams of ordinary light but to the joint quantum properties of entangled photon pairs. The result is a kind of temporal focusing. Quantum correlations, previously smeared across a noisy background, get redistributed into tight, brief peaks, while the incoherent noise remains spread out and can mostly be discarded. Nearly tenfold improvement in signal quality, using off-the-shelf telecom components. No exotic infrastructure required.
When quantum systems send information using individual photons, those photons often travel alongside vast amounts of unwanted light from sunlight, optical amplifiers, or other sources. Because quantum states cannot be amplified or copied without being destroyed, you can’t simply boost the signal to rise above the noise the way classical communication systems do. The useful quantum photon ends up buried and, in practical terms, unreadable.
The device applies carefully designed phase shifts to incoming photons in both the time and frequency domains, effectively acting like a lens array in time rather than space. Because entangled photons are correlated, they respond to these manipulations coherently and focus into narrow temporal peaks. Noise photons lack this coordination and stay spread out. By looking only at what arrives within those peaks, the system discards most of the noise without touching the quantum state carried by the signal photons.
Most quantum photonics experiments rely on specialised, expensive, or hard-to-scale equipment. Running this method at 1550 nanometres in the standard telecommunications C-band means it is compatible with existing fibre infrastructure and commercially available modulators and detectors. That’s a meaningful step toward eventually deploying quantum communication links over networks that already exist rather than building entirely new dedicated infrastructure.
That is one of the applications the researchers have in mind. Satellite quantum links face particularly severe noise problems because the signal photons must pass through daylight or scattered sunlight, which injects enormous amounts of stray light. The INRS team tested the method against noise conditions designed to mimic broadband optical noise of the kind you’d encounter in free-space channels, and the performance improvements held up, though a demonstration on an actual satellite link remains a future challenge.
Quantum state tomography is a technique for fully reconstructing the state of a quantum system by making many carefully chosen measurements on identically prepared copies of it. It gives you the complete statistical description of the state, including how entangled it is and how much it has degraded. In this experiment, the fidelity of a noisy qubit state fell to 0.62 before denoising and recovered to 0.86 after the device processed the photon pairs, a substantial recovery of the lost quantum information.
Quantum states are famously delicate. Unlike classical signals, they can’t simply be amplified if they get weak, and measuring them changes them, which means the usual toolkit for fighting noise doesn’t really apply. You’re stuck with what you’ve got.
The key insight, and it is a rather elegant one, is that coherence itself can do the work. The Talbot Array Illuminator exploits a difference between coherent and incoherent light, a difference that has been understood for decades in the context of spatial optics (where it helps process images), but which the INRS team realised could be extended into the time domain and applied to quantum correlations. Because entangled photons are, by definition, correlated, they respond to the device’s phase manipulations in a coordinated way, focusing into narrow temporal peaks. Random noise photons, lacking that coordination, fail to focus and stay spread out. Postselection, the straightforward business of looking only within the peaks, then filters out the bulk of the noise without touching the quantum information.
“With this new methodology, we were able to recover quantum states corrupted by large amounts of noise, states that would otherwise have been lost,” said Benjamin Crockett, who led the work during his PhD at INRS and is now a Banting postdoctoral fellow at the University of British Columbia. “This could allow quantum systems to operate under real-world noise conditions, helping overcome one of the major barriers to the practical deployment of quantum technologies.”
The numbers are notable. Under moderate noise injection, the coincidence-to-accidental ratio, essentially the signal’s legibility, jumped from 2.2 to 21.3 after the device processed the photon pairs. Quantum interference visibility improved by nearly 50 percent in some conditions. In tomography measurements designed to capture the full state of a qubit, fidelity under heavy noise fell to 0.62 without the denoising module and recovered to 0.86 with it. Entanglement, which had been degraded to the point of being invisible, came back. All of this ran at 1550 nanometres, right in the standard telecommunications C-band, using the sort of fibre and modulator components you might find in any well-equipped photonics lab.
What makes this more than a neat proof of concept is the generality of it. The technique does not require detailed knowledge of where in time the signal photons are likely to arrive, nor their central frequency, which is the sort of prior information that conventional filtering approaches demand. It works on single photons and on entangled pairs. It can, in principle, be extended to microwave frequencies, to free-space channels, even to acoustic systems. The phase modulation required is bounded to about pi radians on each photon, which puts it well within the capabilities of standard electro-optic technology.
There are limits, of course. The current setup has insertion losses of around 5 dB (though that could probably be halved with better components), and the temporal resolution is constrained by the timing jitter of the superconducting nanowire single-photon detectors used in the experiment. Some noise contributions sharing the exact same time-frequency profile as the signal, certain multiphoton artefacts from the source, cannot be separated by this method. And the team has only demonstrated the technique in a relatively controlled laboratory context, not across a deployed city-scale network.
Still, José Azaña, the INRS professor who supervised the work, seemed struck by how clearly it had worked. “Seeing quantum properties emerge in a bright environment, without complex processing steps, was one of the most striking results,” he said.
The next steps the team has outlined are integration onto a chip and testing in real optical fibres and free-space links, perhaps in combination with other denoising methods to push range and reliability further. Chip integration matters because quantum photonic networks, if they are ever to scale, will need components that are compact and manufacturable, not just impressive on an optical bench. Whether this particular technique survives that transition intact remains to be seen, but the underlying principle, that you can exploit quantum coherence as a filter without amplifying or measuring the state, seems robust. It addresses a problem that has stymied quantum networking researchers for years, not by making the channels cleaner, but by making the signals cleverer about surviving dirty ones.
Quantum networks have been perpetually five to ten years away for a while now. Methods like this one won’t change that overnight. But they do, quietly, make the end point feel a little more reachable.
DOI / Source: https://www.science.org/doi/10.1126/sciadv.ady8981
ScienceBlog.com has no paywalls, no sponsored content, and no agenda beyond getting the science right. Every story here is written to inform, not to impress an advertiser or push a point of view.
Good science journalism takes time — reading the papers, checking the claims, finding researchers who can put findings in context. We do that work because we think it matters.
If you find this site useful, consider supporting it with a donation. Even a few dollars a month helps keep the coverage independent and free for everyone.