Shandong team certifies six entangled links in a four-user network with untrusted detectors
Shandong University researchers verified entanglement on all six links of a four-user quantum network without trusting the measurement devices. The method, in Physical Review Letters, drops the extra test photons an earlier approach needed.
The Scientist · Science desk
Bar comparison of links to certify in a fully connected network: four users give 6 links, ten users give 45 links.
Pairs of users to certify, fully connected network In links
| Item | Value | Claim |
|---|---|---|
| Four users | 6 links | 15 |
| Ten users | 45 links | 15 |
What happened
- The network ran on a small photonic chip that makes entangled photon pairs across many wavelengths, with each wavelength channel letting a pair of users share entanglement over 20 km of fiber.
- Entanglement was carried by the possible arrival times of photons, while the orientation of their light waves was used to prepare the known states that test it.
- The team recorded coincidence counts between pairs of users and, from them, set a lower bound on the entanglement shared by each pair.
Why it matters
- exposure A network certified with a conventional test cannot exclude a manipulated detector as the source of its apparent entanglement.
- constraint The remaining trust sits in the preparation of the known test states, so the result holds only if that hardware behaves as specified.
- capability Dropping the ancillary photons removes the extra preparation step that complicated earlier device-independent verification.
- cost Every user added to a fully connected network adds a pair to every existing user, so the links to certify grow from 6 at four users to 45 at ten.
Lu, the paper's senior author, put the problem as one of knowing. "We also need to know whether the entanglement distributed across the network is genuinely there," he told Phys.org [11]. He has worked on it since 2014, when he took part in one of the first experimental demonstrations of measurement-device-independent entanglement witnessing [12].
That earlier test needed additional photons prepared in known states, and the extra photons complicated any real implementation [7]. Lu said his group's work on broadband entangled-photon sources built on integrated lithium-niobate photonics, together with the fast progress of fully connected networks, led them to revisit the idea [13]. The redesign puts both jobs on the same photon. "By performing a joint measurement on these two properties of each photon, we could verify entanglement without trusting the measurement devices or requiring additional ancillary photons," Lu said [10].
Lu's earlier work showed that manipulated measurement results could make a conventional test report entanglement that was not there [6]. A check that does not trust the detectors [4] is not exposed to that manipulation.
Phys.org's account says the method still requires the known test states to be prepared reliably [9]. In our view the result narrows the trust base to the preparation side of the apparatus and does not remove it.
The team established a lower bound on the entanglement shared by every pair [2], and that bound is the useful output. The report does not give its values, the coincidence rates or the measurement time. The tests were described as initial [3], on a four-user network with 20 km links [1]. We would want those numbers before reading this as more than a laboratory demonstration of the principle.
What to watch
- Whether the group publishes per-link values for the entanglement lower bounds, along with coincidence rates and measurement times.
- Whether the reliable preparation of the known test states, the remaining trust assumption, can itself be checked independently.
- Whether the same chip and method are run with more than four users, where the number of pairs to certify climbs quickly.
Clarity's read
What the record supports and how the coverage leans. The claims behind it follow.
Reality
- Evidence58
- Adoption
- Insufficient
- Hype gap+10
- Incentives
- Insufficient
- Confidence60
Claim ledger
Ranked by verification strength, evidence, and original report placement.
- [1]
The network was built on a small photonic chip that generates entangled photon pairs over a broad range of wavelengths; distributing pairs through different wavelength channels connected four users so every pair could share entanglement, giving six connections each spanning 20 km of optical fiber.
- [2]
The team recorded photon coincidence counts between pairs of users, verified entanglement in all six connections and established a lower bound on the entanglement shared between each pair, without relying on trusted measurement devices.
- [3]
The researchers confirmed entanglement across the six links in what Phys.org describes as initial tests.
- [4]
The Shandong University approach to verifying entanglement in a quantum network does not require trusting the equipment used to analyze and detect photons, as most previously developed strategies do.
- [5]
The method is outlined in a paper published in Physical Review Letters and was used to verify entanglement in a quantum network with four users.
- [6]
In earlier work, Lu and colleagues showed that manipulating measurement results could trick conventional entanglement tests into detecting entanglement when none was present.
- [7]
Measurement-device-independent entanglement witnessing overcame that limitation, but it required preparing additional photons in known states for testing, which complicated its implementation.
- [8]
The method uses the possible arrival times of photons to carry entanglement and the orientation of the photons' light waves to prepare known states for testing it.
- [9]
The method still requires the known test states to be prepared reliably.
- [10]
"By performing a joint measurement on these two properties of each photon, we could verify entanglement without trusting the measurement devices or requiring additional ancillary photons," Lu said.
- [11]
"We also need to know whether the entanglement distributed across the network is genuinely there," Lu told Phys.org.
- [12]
Lu took part in 2014 in one of the first experimental demonstrations of measurement-device-independent entanglement witnessing.
- [13]
Lu's group has been developing broadband entangled-photon sources based on integrated lithium-niobate photonics, and with fully connected quantum networks advancing rapidly, these developments motivated revisiting measurement-device-independent verification in networks.
- [14]
Four users with every pair connected give six links, matching the six connections reported.
- [15]
The number of pairs to certify in a fully connected network is n(n-1)/2, which is 6 for four users and 45 for ten users.
Get this ledger inside ChatGPT. Ask about any of our stories and the answer brings its figure, each claim’s verdict and the sources. Free, no account.
Sources
1 independent publisher whose own reporting we read for this story.
- phys.orgQuantum network test checks entanglement without having to trust detectors
1 article · October 14, 2026
Topics and entities
Follow any of these and your For You feed starts watching them — no settings page required.
Topics
- Entanglement verificationFollow
- Quantum networksFollow
- Integrated photonicsFollow