Satellites promise global-scale quantum networks

TL;DR

Photonic loss challenges in fiber and free-space links motivate a spectrum of approaches to a quantum internet. The article surveys ground-based quantum repeaters, memory requirements, and all-photonic concepts, then focuses on satellite-enabled strategies that circumvent exponential attenuation, highlighted by the Micius program. It expands on memory-based satellite protocols and novel satellite-relay architectures (ASQN) that minimize diffraction losses, while contrasting ground-based vacuum beam-guides as an alternative. The paper argues for a staged, hybrid deployment—combining memory-assisted satellites, memoryless satellite relays, and ground networks—to achieve scalable, global quantum entanglement distribution and secure communication.

Abstract

Academia, governments, and industry around the world are on a quest to build long-distance quantum communication networks for a future quantum internet. Using air and fiber channels, quantum communication quickly faced the daunting challenge of exponential photon loss with distance. Quantum repeaters were invented to solve the loss problem by probabilistically establishing entanglement over short distances and using quantum memories to synchronize the teleportation of such entanglement to long distances. However, due to imperfections and complexities of quantum memories, ground-based proof-of-concept repeater demonstrations have been restricted to metropolitan-scale distances. In contrast, direct photon transmission from satellites through empty space faces almost no exponential absorption loss and only quadratic beam divergence loss. A single satellite successfully distributed entanglement over more than 1,200 km. It is becoming increasingly clear that quantum communication over large intercontinental distances (e.g. 4,000-20,000 km) will likely employ a satellite-based architecture. This could involve quantum memories and repeater protocols in satellites, or memory-less satellite-chains through which photons are simply reflected, or some combination thereof. Rapid advancements in the space launch and classical satellite communications industry provide a strong tailwind for satellite quantum communication, promising economical and easier deployment of quantum communication satellites.

Figures (7)

• Figure 1: (a) Teleportation - Initially distant qubits 2 and 3 are entangled (shown by solid blue line), while qubit 1 is separate with its own quantum state but physically close qubit 2. In the next step, the qubit state of 1 is teleported to the distant qubit 3 by doing a joint operation on qubits 1 and 2. This operation (shown by green dashed circle) consists of an entangling gate followed by measurement. For photonics qubits this can be Bell-state measurement, while for atomic qubits it is a two-qubit entangling gate followed by measurement. Further classical communication of the measurement results and single qubits gates on qubit 3 based on the measurement results is needed to complete the teleportation process. This is not shown in the figure for simplicity and mentioned in detail in text. (b) Teleporting an entangled qubit (say II here) to a distant location (say, at the place of IV) forms the basis of a quantum repeater. This enables entangled distribution over longer distances. (c) A schematic diagram showing the working principle of quantum repeaters. Initially, eight quantum memories (A-H) are not connected to each other. Eventually, entanglement between A and H was established though successive tries of entanglement generation, storing and eventually performing the entanglement swapping operation (See text for details). The figure is reproduced from goswami2021photonic with permission.
• Figure 2: Different experiments performed with the Micius satellite yin_satellite--ground_2017yin_satellite-based_2017ren_ground--satellite_2017liao_satellite--ground_2017, as described in goswami2021photonic. (a) Downlink QKD - Using a weak coherent pulse (WCP) source (black circle) photons are sent downlink to ground station to perform decoy-state QKD liao_satellite--ground_2017. (b) Entanglement based QKD - An entangled pair source (red circle) aboard Micius is used to perform QKD between the satellite and the ground station. (c) Uplink teleportation - Entangled photon pair source (red circle) in ground station is used to teleport a qubit to the satellite. The unknown qubit, to be teleported, also comes from another entangled pair for technical reasons. See text for details ren_ground--satellite_2017. (d) Entanglement distribution in downlink- Entangled photon pairs are distributed between two ground stations, separated by a record 1203 km on earth, by double downlink transmission yin_satellite-based_2017.
• Figure 3: Two repeater protocols over LEO satellites to distribute entanglement over global distances using quantum memories (QM) and quantum non-demolition (QND) detectors in ground stations boone_entanglement_2015. In (a) the QM and QND detectors are kept in ground stations for operation advantages while in (b) the QM and QND detectors in the intermediate links (dashed box in (a)) are kept in the satellite to avoid extra intermediate ground link connections gundogan_proposal_2021liorni2021quantum.
• Figure 4: Entanglement is distributed using very long storage-time memory in a single satellite. (a) One Photon of an entangled pair is stored in the memory (red Square) which is later retrieved and transmitted down-link once the satellite has physically moved to a far-away destination simon2017towardswittig2017concept. (b) Rates of the single-satellite protocol can be dramatically improved by using a time-delayed repeater protocol using two memories, that result in complete usage of the memory multiplexing capacity gundogan2024time.
• Figure 5: The satellite relay architecture of ASQN is shown, for global scale quantum communication without quantum memory or quantum repeaters. (a) Photons are sent from one side of the globe to another by being reflected from one satellite to another and being guided along the surface of the earth (satellite separation exaggerated). (b) Each satellite has telescopes with curved mirrors that together focuses the light, while also bending it in the direction of the next satellite. Effectively the whole telescope mirror assembly in each satellite behaves like one lens (as shown), without considering the light bending that can't be modeled by a lens. (c) The chain of satellites behave like a set of lenses and contain light beam divergence or diffraction loss indefinitely over very long distance. Diffraction loss at 20,000 km is nearly eliminated (0.67 dB only), while total loss is contained below 30 dB - for satellite telescope diameter of 60 cm, satellite separation of 120 km and 2$\%$ reflection loss at each satellite goswami2023satellite.