Building a Quantum Network with “Satellite Lenses”

Researchers and industry are increasingly eyeing the prospect of global communications networks that would take advantage of the security offered by quantum technology. One stumbling block, though, has been the lack of scalable “quantum repeaters” analogous to the ones that keep optical signals alive in long-haul classical fiber networks.

As an alternative, some research groups are looking at satellite-based quantum communications, in which quantum information would ride on laser beams between spacecraft in low Earth orbit (LEO). Yet even satellite schemes have their pitfalls. Loss of photons in diffracting laser beams, as well as the curvature of the Earth itself, would likely limit realistic distances of high-efficiency quantum links between LEO satellites to less than 2000 km.

Now, researchers Sumit Goswami of the University of Calgary, Canada, and Sayandip Dhara of the University of Central Florida, USA, have laid out a proposal showing how those pitfalls could be overcome (Phys. Rev. Appl., doi: 10.1103/PhysRevApplied.20.024048). Their proposal involves relaying delicate quantum signals across a chain of relatively closely spaced, synchronously moving satellites. Those satellites, the pair suggests, could effectively act “like a set of lenses on an optical table,” focusing and bending beams along Earth’s curvature and preventing photon loss across distances as great as 20,000 km—without the need for quantum repeaters.

It’s all done with mirrors

While Goswami and Dhara metaphorically refer to the nodes in their proposed all-satellite quantum network (ASQN) as satellite lenses, in reality the optical magic happens with mirrors, to keep absorption-related photon losses to an absolute minimum. In simplified terms, a given satellite in the chain sends a beam of light to the next one, perhaps 120 km away. That next satellite captures and refocuses the beam with a receiving mirror and bounces it off of two smaller mirrors to a final transmitting mirror, which relays the signal on to the next satellite in the chain.

Under their proposal, the researchers say, closely spaced satellites effectively act “like a set of lenses on an optical table,” focusing and bending beams along Earth’s curvature and preventing photon loss due to diffraction.

In their modeling, Goswami and Dhara considered a chain of satellites, each separated from the next by 120 km; given expected beam divergence in Earth orbit, that implies a telescope diameter of 60 cm for each satellite. The team’s modeling suggests that such a relay setup, with the quantum signal passed from satellite to satellite by reflection, would virtually eliminate diffraction loss across distances of 20,000 km.

Managing other losses

With diffraction loss taken care of, Goswami and Dhara methodically looked at other potential sources of loss in the satellite-lens system. One obvious one is reflection loss of some photons at the mirrors themselves, which the pair thinks could be kept manageable through a configuration combining large metal mirrors and small, ultrahigh-reflectivity Bragg mirrors. Another source of loss lies in tracking and positioning errors for the satellites in the chain; such hiccups would need to be held to a minimum to keep the satellites in sync with one another.

A final source of loss has nothing to do with the satellites. Depending on the quantum communication architecture, quantum information needs to be transmitted from and to stations on Earth’s surface. For free-space optical signals, that opens the prospect of data losses due to atmospheric turbulence, which can dramatically increase the beam size and spread.

Turbulence turns out to be a much bigger problem for data in the uplink (ground to satellite) than in the downlink (satellite to ground). That’s because in the uplink, the turbulence is doing its dirty work at the beginning of the communication chain rather than at its end, and the turbulence-induced beam divergence and fragmentation are magnified as the beam propagates. In the system proposed by Goswami and Dhara, however, focusing of the light captured at the first satellite after the uplink prevents the turbulence effect from being magnified across the long satellite chain as a whole.

Outperforming fiber—without repeaters

Taking all of these sources of loss (and a few others) into account, Goswami and Dhara numerically simulated how such a chain of relay satellite lenses might work in transmitting quantum information under two scenarios. One is so-called entanglement distribution, the protocol demonstrated by researchers in China on the Micius satellite, in which photons are entangled in space and sent in different directions via the satellite lenses, ultimately to be transmitted down to widely separated stations on Earth and tested for quantum security.

The other is a simpler “qubit transmission” protocol, in which quantum bits (qubits) are simply sent from a ground station to the first satellite, transmitted across the chain and finally beamed down to a second, distant ground station. Such a system would require a different kind of optical design, to counteract the impact of turbulence on the satellite uplink. Goswami and Dhara think this approach may have certain advantages, however, as it keeps both the qubit source and detection in more controllable, better-outfitted ground stations.

Under the entanglement distribution scenario, the team found that the total signal loss across 20,000 km would come in at around 30 dB. That’s comparable to the loss experienced across only 200 km of a direct optical-fiber link, assuming a loss rate of 0.15 dB/km in the fiber. (The loss for qubit transmission, including loss on uplink turbulence, was a higher 50 dB across 20,000 km, assuming satellites orbiting at an elevation of 500 km.) “Such a low-loss satellite-based optical-relay protocol,” Goswami and Dhara write, “would enable robust, multimode global quantum communication and would not require either quantum memories or repeater protocol.”

Substantial engineering needed

“What this proposal basically does,” Goswami observed in an email to OPN, “is that it shifts the task of creating quantum network from physics to engineering.” He added, however, that some of the engineering likely wouldn’t be trivial, particularly with respect to designing and developing the satellites in the fleet. Still, he and Dhara stress in the paper that recent developments in space technology—embodied in reusable launch vehicles from organizations such as SpaceX and the vast constellations of classical-communications satellites being lofted into LEO by a number of private companies—make a system such as their ASQN considerably more feasible than it would have been in the past.

Goswami and Dhara stress that recent developments in space technology make a system such as their ASQN considerably more feasible than it would have been in the past.

Goswami told OPN that a chain of around 160 satellites would be needed to cover the full 20,000-km distance modeled in the paper. Such a single chain, he noted, would cover most of the globe every three days—so, Goswami said, “even just one chain can be used for connecting many places at different times.” But a larger, 2D network, to enable uninterrupted worldwide quantum communications, would require tens of thousands of new satellites.

Quantum network yes, quantum internet maybe

Goswami and Dhara believe that, by dispensing with the need for quantum repeaters or memory, the scheme they’ve proposed and modeled could open a range of possibilities implicit in a quantum network. Such prospects include secure communication via quantum key distribution and precision long-distance quantum sensing.

The researchers admit, however, that a more complex network—that is, the long-term vision of a “quantum internet” now being fleshed out in a variety of research labs—would still require some sort of quantum memory to ensure completely lossless transmission. Still, Goswami and Dhara argue that, by eliminating diffraction loss, their setup would relax some of the more stringent efficiency requirements of the needed quantum memory. Thus certain configurations of their ASQN, they write, could not only serve to build an orbiting quantum network, but could prove “another interesting candidate for implementing the quantum internet.”

Corrections: On 28 August 2023, at 07:45 EDT, the story was updated to clarify several model details, and to note the difference in calculated total loss between the two quantum communication scenarios tested. On 29 August at 10:30 EDT, the story was updated to clarify that in the proposed setup, loss attributable to turbulence after uplink does not propagate after the first satellite.