Tampere University and Nokia Bell Labs Adapt MIMO Space-Time Coding to Rydberg Atomic Quantum Receivers in a September 2026 Paper
Quentir Defense Monitor
Evidence-based insights for quantum defense and security. Published by Quentir Systems LLC · September 4, 2026.
A quantum radio receiver made of atoms has no antenna, and until now it also had no good way to work inside the multi-antenna architectures that modern wireless links depend on. On September 1, 2026, a team of six researchers from Tampere University and Nokia Bell Labs in Finland posted a five-page paper to arXiv that addresses exactly that gap. Asifa Zannat, Milad Abolpour, Dani Korpi, Mikko A. Uusitalo, Mikko Valkama, and Ertugrul Basar propose a space-time coding architecture that lets Rydberg atomic quantum receivers deliver the full diversity gain of a multiple-input multiple-output link, the property that makes a radio connection survive fading and interference. The paper is under review at IEEE, and its results come from analysis and numerical simulation at a 5 GHz carrier.
The problem the team solves is baked into the physics of the sensor. A Rydberg receiver detects radio waves with a vapor of atoms pumped by lasers into highly excited states, where the outermost electron orbits so far from the nucleus that the atom responds to faint electric fields the way a large antenna would. The readout, though, is optical: the radio field shifts the atomic energy levels, and a probe laser measures the size of that shift. What comes out is the magnitude of the field. The phase, the timing information that conventional radio engineering leans on for almost everything, is lost in the measurement. Every multi-antenna technique in the standard toolbox, from beamforming to spatial multiplexing, assumes a receiver that captures both amplitude and phase as a complex number. An atomic sensor that reports only magnitude breaks that assumption at the first line of the signal model.
For a defense reader this is a detail with a program attached. Rydberg sensors are the leading candidate for a new class of receive apertures, and several national-security organizations have spent the past years proving the front end works. The receiver-side signal processing that turns a working sensor into a working link has lagged behind, and the Finnish paper is a concrete step on that second track.
How a strong reference field and real orthogonal codes give the atomic sensor a linear MIMO model
The architecture rests on two moves. The first is a known trick from atomic sensing: inject a strong, known reference radio signal at the receiver alongside whatever the link delivers. The reference acts like the local oscillator in a classical heterodyne radio. Because the atoms respond to the combined field, the magnitude they report now rises and falls with the phase relationship between the reference and the incoming signal, and phase information becomes recoverable from a magnitude-only readout. The paper states the operating condition plainly: the reference must dominate the received signal and the noise at every atomic sensing element for the linearization to hold.
The second move is the coding choice. Instead of sending independent data streams from each transmit antenna, the transmitter repeats symbols across antennas and time slots according to a real orthogonal design, a structure from the classical space-time coding literature chosen here because it survives the constraints of the atomic readout. The orthogonality means the receiver can pull the symbols apart with a simple matched filter, with no matrix inversion and no iterative detector. The authors derive the equivalent linear model, prove the decoupling, and give closed-form bit error probability expressions. In their simulations, the scheme achieves the full diversity order, the product of transmit antennas and atomic receive elements, and at high signal-to-noise ratio it outperforms the spatial multiplexing baselines they compare against. Diversity order describes how steeply the error rate falls as signal quality improves; full diversity is the ceiling a given antenna count allows.
What the paper does under honest assumptions is worth stating with equal clarity. The channel model is standard Rayleigh fading with Gaussian noise, the reference field is assumed perfectly known, and the atomic response is treated through the linearized model throughout. Nothing was run on hardware. The contribution is a receiver architecture and its mathematics, worked out to the point where an experimental group could test it on a real vapor cell.
Quantum pillar: sensing (RF spectrum sensing). Use posture: dual-use. Technology readiness: TRL 2 of 9. The coding architecture exists as mathematics and numerical simulation of a 5 GHz link, with closed-form error expressions in place of any run on a physical atomic receiver, so a buyer should read it as a concept whose hardware trial still lies ahead.
Where this sits in the defense pipeline: DARPA Quantum Apertures, an Army 0 to 20 GHz sensor, and Infleqtion field trials
The reason a coding paper from a telecom group belongs in a defense briefing is the decade of government work that precedes it. DARPA's Quantum Apertures program set out to build radio-frequency receivers from Rydberg atoms precisely because they promise sensitivity and frequency agility that conventional front ends cannot match, and the agency listed electromagnetic spectrum operations, radar, and communications as the national-security areas at stake. The program has since concluded, and its performers included the industrial teams that now carry the technology forward. The U.S. Army's research laboratory built a waveguide-coupled Rydberg spectrum analyzer that samples from zero frequency to 20 GHz with a single atomic sensor, a span that would take a rack of conventional antennas and amplifiers.
Metrology and productization have kept pace. NIST researchers demonstrated reception of a consumer handheld UHF two-way radio with a Rydberg sensor, published in Physical Review Applied in February 2026, resolving neighboring channels with 53 decibels of isolation using the same offset-local-oscillator idea the Finnish paper builds on. On the commercial side, Infleqtion sells a Rydberg-based quantum spectrum sensing receiver covering 1 MHz to 12 GHz in its second generation, has run it at the Army's C5ISR NetModX assessment, and works with L3Harris on quantum RF sensing for defense platforms. The front end, in other words, has left the laboratory.
Read as capability, the Finnish contribution targets the layer above all of that hardware. RF spectrum sensing with atoms gives a force a receive aperture with no metallic antenna to resonate, detect, or jam, wideband coverage from a single physical package, and a measurement process that absorbs almost none of the energy it observes, which makes the listening hard to notice. Those properties serve signals intelligence and spectrum awareness directly. What they have lacked is the link-layer machinery that makes a receiver into a communications terminal: the ability to combine multiple atomic elements coherently, ride out fading, and hold a data link in a contested environment. Full-diversity reception is exactly that machinery. A force that fields it gains tactical receive terminals that keep working where conventional antennas are being jammed or would give a position away; an adversary facing it loses some of the detectability and denial options that antenna physics used to offer.
The posture is dual-use on its face. The authors are a university group and a commercial network vendor, the target application is wireless infrastructure, and every result in the paper serves a civilian 6G receiver as well as it serves a military one. The same linearization and the same codes would sit in a cell site or in an intercept terminal, and nothing in the mathematics prefers either buyer.
What separates a 5 GHz simulation from a receiver a program office can rely on
The distance between this paper and a fieldable terminal is the ordinary distance between TRL 2 and TRL 7, and it runs through specific, checkable steps. The scheme needs a hardware demonstration on a real multi-element vapor-cell receiver, where laser stability, cell-to-cell variation, and the true atomic noise statistics replace the Gaussian assumptions of the simulation. The strong-reference condition needs engineering: injecting a dominant, phase-stable reference field at every sensing element is straightforward on a bench and harder on a moving platform. Dynamic range questions follow, since a reference strong enough to linearize the readout also sits inside the same atomic response that must still register a weak incoming signal. And the comparison that matters to a buyer, performance per watt and per liter against a mature conventional MIMO front end, does not exist yet for any Rydberg link, coded or otherwise.
None of that is a criticism of the work; it is the map of what remains. Five pages of correct receiver mathematics is how the gap between a sensor demonstration and a communications capability actually closes, one assumption at a time. The groups best positioned to run the next step are the ones already holding hardware: the DARPA program alumni, the Army laboratory with its 20 GHz analyzer, and vendors like Infleqtion with multi-element roadmaps. A program office watching this space should log the paper as an early signal that the telecom industry has started spending its own signal-processing talent on atomic receivers, and should ask its quantum sensing suppliers a concrete question at the next review: when a multi-element atomic receiver arrives, which coding architecture will it run, and has anyone tested it beyond simulation.
Sources
Primary source: Asifa Zannat, Milad Abolpour, Dani Korpi, Mikko A. Uusitalo, Mikko Valkama, and Ertugrul Basar (Tampere University and Nokia Bell Labs), 'A Novel Space-Time Coding Architecture for Rydberg Atomic Quantum Receiver-Based Systems,' arXiv, September 1, 2026. Other material: DARPA's Quantum Apertures program page; NIST's Physical Review Applied paper on Rydberg reception of a handheld UHF radio; Infleqtion's quantum spectrum sensing product documentation.