Can one compact photon detector carry gigabit laser links from space?
A four-output superconducting nanowire detector counted single photons fast and efficiently enough to receive laser data at over a gigabit per second using only a couple of photons per bit, while tolerating daylight-level noise.
Source
A compact multi-pixel superconducting nanowire single-photon detector array supporting gigabit space-to-ground communications
Study at a glance
- Design
- Other — Lab device experiment: a 2×2 NbN nanowire detector array with six shunted sub-pixels per quadrant, characterised optically and used as the receiver in a free-space 1550 nm PPM communication testbed.
- N
- One detector device; no sample of units or participants.
- Population
- A single NbN superconducting nanowire single-photon detector array operated at cryogenic temperature
- Outcome
- Detection efficiency, timing jitter, counting rate, photon-number resolution, receiver sensitivity (photons per bit), tolerance to background noise, and beam-position sensing
Structured fields used in claim comparison tables when every cited study has a complete layer.
What they did
The authors built a superconducting nanowire detector split into four quadrants, each made of six nanowire sections in series with shunt resistors, so that many sub-pixels share one readout line. They measured its efficiency, timing jitter, counting rate and ability to count how many photons arrived. They then used it as the receiver in a lab laser-communication testbed with pulse-position modulation and error-correcting codes, varying bit rate, added background light and beam position.
What they found
Average detection efficiency was 91.6%, jitter was 78 ps for single photons and 21 ps for six-photon events, and the whole array counted about 1.6 billion photons per second while resolving up to 24 photons. The receiver needed 1.47 photons per bit at 480 Mbps and 7.41 photons per bit at 1.5 Gbps. Using photon-number information let the link tolerate roughly 0.8 background photons per slot at 120 Mbps versus about 0.1 for simple on/off detection, and the four quadrants could sense beam position.
The limits
What it doesn't show
Everything was done in a laboratory with an attenuator simulating distance, not over a real space-to-ground link with atmospheric turbulence. Signal processing and decoding were done offline on recorded waveforms, so real-time operation is not demonstrated. With free-space coupling the system efficiency fell to 52.5%, and the small active area creates a trade-off with collecting light from a large telescope. Only a single device was tested.
Key terms
- Superconducting nanowire single-photon detector (SNSPD)
- A thin superconducting wire biased just below its critical current, which briefly becomes resistive when it absorbs a single photon, producing a voltage pulse.
- Timing jitter
- The spread in the measured arrival time of a detected photon, which limits how finely time slots can be divided.
- Pulse-position modulation (PPM)
- Encoding data by which time slot within a symbol contains a light pulse; it is energy-efficient for photon-starved links.
- Photon-number resolution
- The ability of a detector to report how many photons arrived at once, not just whether any arrived.
- Kinetic inductance
- Inductance arising from the inertia of superconducting charge carriers; longer nanowires have more, which slows their recovery after a detection.
Flashcards
0 of 11 answers reviewed
Research intelligence for this paper
See its role on concept claims, tensions it is part of, placement history, and related discoveries.
Quiz yourself
What allows the detector to count photons quickly despite a large total wire length?
Common questions
Why split the detector into many series sub-pixels instead of one long wire?
A long wire has high kinetic inductance and recovers slowly; shunted short sections let unfired sections keep detecting while one recovers, raising the counting rate without needing a separate readout line per pixel.
How does counting photon numbers help against daylight noise?
The error-correction decoder can weigh slots by how many photons they contain, so a slot with several signal photons is distinguished from one containing a stray background photon, raising the tolerable noise level.
What is the cost of putting sub-pixels in series?
The output pulse gets smaller, reducing signal-to-noise and worsening timing jitter, which the authors offset with cryogenic amplifiers.
More on Superconductivity