USTC Demonstrates Quantum Position Verification With 75-Meter Accuracy Over 2 Kilometers of Fiber, Published in Nature Physics on September 3, 2026
Quentir Defense Monitor
Evidence-based insights for quantum defense and security. Published by Quentir Systems LLC · September 13, 2026.
On September 3, 2026, Nature Physics published the first complete experimental demonstration of quantum position verification, the work of Guang-Can Guo's quantum information laboratory at the University of Science and Technology of China in Hefei together with the Institute of Advanced Photonics Technology at Guangdong University of Technology. Two verifier stations separated by two kilometers of fiber interrogated a prover with faint pulses of light and confirmed where that prover actually stood to better than 75 meters, with the security of the answer resting on quantum physics and the speed of light rather than on any trust in the prover's own equipment. The paper, from first author Guan-Jie Fan-Yuan and twelve colleagues under the direction of Zheng-Fu Han, closes a gap that has stood open since theorists first proposed position-based cryptography: every earlier scheme either lacked a security proof or demanded hardware performance nobody could build.
Position is quietly load-bearing across military and civil systems, and almost everywhere it is taken on faith. A ship reports its coordinates, an aircraft transponder broadcasts a track, a logistics tag claims a warehouse, and the receiving system believes the claim because it has no independent way to check it. Quantum position verification turns the claim into a test. The verifiers do not ask the prover where it is; they measure whether anything at the claimed point responded the way only something at that point could. For a defense reader the interesting part is exactly that inversion, because the growing family of position attacks, from falsified vessel tracks to spoofed satellite navigation, all exploit systems that accept location claims without an independent check.
This briefing reads the experiment as a capability in the making: what the Hefei and Guangzhou teams actually built, why classical physics cannot deliver the same test, and what still separates a laboratory fiber loop from a verification service a program office could buy.
What Fan-Yuan and colleagues built: weak laser pulses, a Sagnac encoder and a 247.8-nanosecond answer window
The protocol works like a timed oral exam held from two directions at once. Each verifier occupies a surveyed point and both are connected to the prover by optical fiber. One verifier sends a quantum state, a faint laser pulse carrying polarization information; the other sends a classical instruction, in the experiment a string of 40 bits, timed so both arrive at the claimed position in the same instant. The prover must combine them, perform the measurement the instruction dictates and return the result to both verifiers immediately. Because information cannot travel faster than light, a response that arrives on time at both stations can only have been computed at the claimed position. An impostor standing anywhere else faces a deadline it cannot meet, and the quantum pulse cannot be copied and forwarded to accomplices, because unknown quantum states resist copying as a matter of physical law.
Turning that logic into hardware is where the Nature Physics paper earns its place. The team dropped the ideal single photons of earlier theory for phase-randomized weak coherent states, ordinary attenuated laser pulses of the kind mature quantum key distribution systems use, and reworked the security analysis so the protocol survives realistic loss. A Sagnac interferometer built around a micro-assembled rotated circulating splitter prepares the polarization states, and it held the quantum bit error rate to 0.27 percent. Latency discipline runs through the whole system, because every nanosecond of processing slack loosens the position bound by about thirty centimeters. The classical signals travel through anti-resonant hollow-core fiber, where light moves through air rather than glass and arrives roughly a third sooner, the measurement bases switch at 2 megahertz, superconducting nanowire detectors catch the pulses with 90 percent efficiency, and a field-programmable gate array answers the 40-bit instruction from a one-terabit lookup memory. The complete chain, from challenge arrival to answer on its way, stays under 247.8 nanoseconds, and that response window is what converts to the 75-meter accuracy, roughly the footprint of a single large building. A summary from the university, carried by China Science Daily's ScienceNet, frames the intended applications the same way the paper does: verifying claimed positions for high-value asset custody, emergency response and position-based access to sensitive systems.
Quantum pillar: sensing (inertial navigation and PNT). Use posture: defensive. Technology readiness: TRL 4 of 9. The full protocol ran as an assembled system across two kilometers of laboratory fiber with every component real, and trials over deployed cable plants, longer ranges and moving platforms all still lie ahead.
Why classical position checks always fall to colluding relays, and what Buhrman's impossibility theorem concedes to entangled adversaries
The reason this took a quantum experiment is that the classical version of the problem is settled, and it is settled against the defender. Any classical challenge, however cleverly randomized, is just bits, and bits can be copied. A ring of colluding adversaries surrounding the claimed position can each intercept the challenge nearest to them, share copies at light speed, compute the expected answer and reply with timing indistinguishable from a genuine prover at the claimed point. No protocol design escapes this, which is why position verification never became a classical security primitive despite decades of interest in distance-bounding radio protocols. Quantum states change the arithmetic because the pulse carrying the challenge cannot be duplicated, so the colluders cannot all possess it at once, and the timing trap closes.
The quantum story then took its own turn. In 2010, Harry Buhrman, Serge Fehr, Christian Schaffner and colleagues proved that adversaries sharing an arbitrarily large stock of entanglement can still defeat any quantum position verification scheme, using a technique that spreads a quantum computation across separated attackers in a single round of communication. The same paper showed the attack needs entanglement in quantities that grow with the protocol's complexity, and that adversaries without pre-shared entanglement can be beaten soundly. Practical security therefore rests on an explicit resource assumption: the protocol is safe against any attacker whose entangled quantum memory falls below a threshold the defender can raise by design choices. No adversary today fields entangled memory at anything near the scale required, and the assumption is checkable against the open state of quantum hardware, which is a firmer footing than the untested hardware assumptions much of deployed security already leans on. The Hefei experiment's contribution sits on the other side of the ledger: it shows the honest parties' half of the protocol, long considered the impractical half because of its loss and latency demands, runs on components that exist.
What a 75-meter trusted fix offers GPS-denied operations, blue-force pictures and asset custody
The operational backdrop is a position-integrity problem that has grown measurably worse. The OPSGROUP working group's final report of September 2024 counted roughly 1,500 civil flights a day experiencing satellite navigation spoofing, a fivefold rise inside a year, concentrated around conflict regions where militaries broadcast false signals. Defense responses so far concentrate on making a platform's own navigation independent of the vulnerable signal, and quantum sensing is prominent in that effort. This publication's review of SandboxAQ's magnetic navigation flight test on Northrop Grumman's Lumberjack drone covered one such program, and the broader survey in Quantum Navigation Leaves the Laboratory Bench maps the field. All of those systems answer the question a platform asks itself: where am I. Position verification answers the question everyone else must ask: is that platform where it says it is. The two capabilities are complements, and the second has had no rigorous answer of any kind until now.
Read as a defense capability, verified position becomes a credential. A command network could accept a blue-force position report only after the reporting node passes a verification round, which would stop an adversary from injecting phantom friendly tracks into a common operating picture. Position-based authentication could bind access to cryptographic material or weapons-release networks to a physically verified location, so a captured terminal moved elsewhere stops working, a control the USTC authors explicitly foresee for financial transactions and sensitive data access. Custody of nuclear material, treaty-limited items or high-value cargo could rest on a check the custodian cannot forge. The gains run to the defender in each case, which is why this briefing reads the posture as defensive: the capability validates and authenticates, and an attacker gains from its absence rather than its presence.
A program office should also see clearly what the experiment does not establish. The demonstration ran on two kilometers of fiber in laboratory conditions, and the paper's own security model covers attackers without large entangled memories, an assumption that will need monitoring as quantum hardware matures. The 75-meter figure suits authentication at the scale of a building or a compound; it does not support targeting-grade fixes, and it verifies cooperative provers only, saying nothing about locating an adversary who declines to answer. The verifier stations themselves need surveyed positions, synchronized timing and cryogenic single-photon detectors, and extending the scheme beyond fiber to free-space links, the step that would bring aircraft and ships into reach, remains undemonstrated. Those are the distances between a physics result and a fielded service. What the result changes today is the answer to a buyer's threshold question: position verification with a security proof now exists as working hardware, and the components it runs on come from the same supply chain the quantum communications industry is already scaling.
Sources
Primary source: Guan-Jie Fan-Yuan and colleagues, Relativistic position verification with coherent states, Nature Physics, September 3, 2026 (USTC Hefei, Guo Guangcan laboratory, with Guangdong University of Technology). Other material: China Science Daily's ScienceNet summary of September 12, 2026; Buhrman et al., Position-Based Quantum Cryptography (2010); the OPSGROUP GPS Spoofing WorkGroup final report (September 2024).