Elias Towe, Carnegie Mellon University, introduced three speakers invited to address quantum communications and networks: Prem Kumar, Northwestern University; Saikat Guha, University of Arizona; and Nitin Samarth, Pennsylvania State University. Alan Willner, University of Southern California, moderated a brief Q&A following each presentation.
Prem Kumar, Northwestern University
Kumar described the differences between quantum communications and classical communications and discussed the path to quantum communications and, ultimately, to quantum networks.
Classical Communications Versus Quantum Communications
In classical communications, error-free communication is possible below a certain capacity. Replacing classical bits with qubits creates conflicts with quantum mechanics, the no-cloning theorem, and the Heisenberg uncertainty principle (the impossibility of measuring a quantum state), making it impossible to directly extend from classical to quantum communication. However, the teleportation protocol showed that a combination of classical communications and quantum
entanglement can alleviate those difficulties and achieve a certain level of quantum communication.1 Similar techniques have been used for continuous variables of quantum information,2,3 but the challenge now is to reliably, and over long distances, distribute quantum-entangled information from one point to another.
Progress Toward Quantum Communication
In classical communications, the development of optical communications via pulses of light or encoded light rays, advanced with the invention of the erbium optical amplifier, was a key enabler of reliable, cost-efficient, long-distance global communications. Could a similarly game-changing advancement occur for quantum communication? Although the solution is far off, necessary components like quantum repeaters and trusted satellites are already being developed. For example, J. Yin et al. reported on perfecting United States–developed technology, launched a trusted satellite, Micius, capable of long-distance quantum communication over 1,200 km.4 The satellite breached the previous distance limit and was one of many firsts that Chinese researchers have achieved. Singapore was able to package simple quantum key distribution and launch it on a CubeSat.5
Kumar’s laboratory showed that quantum communication was possible on classical communications infrastructure, a possibility that could be exploited for various applications.6,7,8 By engineering quantum sources over classical sources,
1 C.H. Bennett, G. Brassard, C. Crépeau, R. Jozsa, A. Peres, and W.K. Wootters, 1993, Teleporting an unknown quantum state via dual classical and Einstein-Podolsky-Rosen channels, Physical Review Letters 70:1895-1899, https://doi.org/10.1103/PhysRevLett.70.1895.
3 A. Furusawa, J.L. Sørensen, S.L. Braunstein, C.A. Fuchs, H.J. Kimble, and E.S. Polzik, 1998, Unconditional quantum teleportation, Science 282(5389):706-709, https://science.sciencemag.org/content/282/5389/706.abstract.
4 J. Yin, Y, Cao, Y.-H. Li, S.-K. Liao, L. Zhang, J.-G. Ren, W.-Q. Cai, et al., 2017, Satellite-based entanglement distribution over 1200 kilometers, Science 365(6343):1140-1144, https://doi.org/10.1126/science.aan3211.
6 M. Fiorentino, P.L. Voss, J.E. Sharping, and P. Kumar, 2002, All-fiber photon-pair source for quantum communications, IEEE Photonics Technology Letters 14(7):983-985.
7 X. Li, P.L. Voss, J.E. Sharping, and P. Kumar, 2005, Optical-fiber source of polarization-entangled photons in the 1550 nm telecom band, Physical Review Letters 94(5):053601, https://doi.org/10.1103/PhysRevLett.94.053601.
8 C. Liang, K.F. Lee, J. Chun, and P. Kumar, 2016, “Distribution of Fiber-Generated Polarization Entangled Photon-Pairs over 100 km of Standard Fiber in OC-192 Environment,” Paper PDP35, Optical Fiber Communication Conference and Exposition and the National Fiber Optic Engineers Conference, Technical Digest (CD), Optical Society of America, Anaheim, Calif., https://www.osapublishing.org/abstract.cfm?uri=OFC-2006-PDP35.
quantum teleportation over a network is now possible, and current technology can be used to demonstrate a rudimentary quantum network.9
Challenges to Creating Quantum Networks
While Kimble imagined a quantum internet a decade ago, many of the components and techniques toward this goal are still being refined.10 However, quantum information processing (QIP) is advancing rapidly, and there are many different quantum computing platforms in development that could work for QIP. Kumar and his colleagues envision a system in which even very different quantum devices and technologies, such as those based on atoms, ions, photons, and superconductors, can be interconnected.
One major challenge to quantum networks is establishing quantum state transduction, where quantum states can transfer from one physical modality to another. It is very difficult to convert with 100 percent efficiently and noiselessly, Kumar said. It is also difficult to convert microwave photons to optical photons, although several groups are attempting to do so. Another key challenge is how to create and control quantum switching and routing in order to connect the various quantum platforms. However, experiments are showing progress in quantum eye opening, which enables routing of qubits while preserving their quantum state, meaning that they can be moved to a specific system or link.11
Concluding, Kumar said that despite the challenges, today’s noisy intermediate-scale quantum (NISQ) era is the right time to begin planning quantum networking. A quantum communications network with interconnecting m n-qubit machines is not far off, in his view. However, there are many technical challenges that need to be overcome, including reliable, on-demand sources of entangled-pair photons; quantum repeaters; and more basic architectural problems such as system layering, control, management, and user access. Kumar recommended the recent U.S. Department of Energy (DOE) workshop report, Quantum Networks for Open Science, for more information.12
9 Q.C. Sun, Y.-L. Mao, S.-J. Chen, W. Zhang, Y.-F. Jiang, Y.-B. Zhang, W.-J. Zhang, et al., 2016, Quantum teleportation with independent sources and prior entanglement distributed over a network, Nature Photonics 10:671-675, https://doi.org/10.1038/nphoton.2016.179.
11 M.A. Hall, J.B. Altepeter, and P. Kumar, 2011, All-optical switching of photonic entanglement, New Journal of Physics 13:105004, https://iopscience.iop.org/article/10.1088/1367-2630/13/10/105004/meta.
12 Office of Advanced Scientific Computing Research, Office of Science, U.S. Department of Energy, 2018, Quantum Networks for Open Science Workshop Report, https://www.orau.gov/quantumnetworks2018/Quantum-Networks-for-Open-Science-Workshop.pdf.
In response to a question, Kumar noted that in his experiments, quantum routers were based on standard optical fiber because it was readily available, but it could also be possible to use a chip-scale format. Any material with a good third order of nonlinearity could work well with a few modifications, he said.
Saikat Guha, University of Arizona
Guha described the field of photonic quantum information processing (QIP) and how photonics can be used for quantum communications and quantum repeaters.
Photonic QIP—at the intersection of quantum optics, quantum information, algorithms, optical imaging and sensing, and network communication theory—aims to harness quantum mechanical properties of light in all applications involving acquisition, encoding, transmission, and processing of information via light. Although the all-photonic route to universal quantum computing might eventually prove a serious contender to other qubit technologies, there are many high-impact nearer-term applications of photonic QIP where the information to be computed is natively in the optical domain, such as in optical communications, sensing, and imaging tasks. Guha’s research focuses on optical quantum communications—harnessing the quantum properties of light to bring advancements in classical and quantum communications, and networking. Guha is part of an interdisciplinary team of experts building a photonic QIP research hub and developing a multidisciplinary quantum curriculum at the University of Arizona.
Optical receivers come in many versions, and each induces a certain kind of noise. The classical communications capacity associated with different receivers may be higher or lower in different regimes of operation. Quantum mechanics sets the fundamental limits on classical communications capacity. Guha’s team is working on developing an optical receiver design that can attain the fundamental quantum limit of classical communications capacity, known as the Holevo limit. Such a receiver must use quantum processing of light, and this is a substantial challenge. In principle, classical communications capacity increases as power increases. However, in optical communications, there is a trade-off between the speed (spectral efficiency) and the photon efficiency (bits encoded per photon).13
Quantum communications involves three main tasks: sending qubits reliably across distances; generating shared entanglement across distances; and generating secret, quantum-secured communication across distances. These tasks are closely related, and their maximum rate is limited by the loss in the channel—that is, they cannot be circumvented by increasing transmit power, a peculiar feature of quantum communications. This results in the rate of all three of these tasks diminishing exponentially with the distance—for example, in fiber-based quantum communications. In order to realize quantum communications systems whose rate-versus-distance performance is not limited as above, one will need to build quantum repeaters and intersperse them along the length of the communications channel. Quantum repeaters are special-purpose quantum processors.
Reliable classical communications requires error correction. For quantum communications, similar one-way error correction is necessary—that is, encoding and decoding qubits. However, there is an alternative for quantum communications: establishing entanglement across distances followed by a process known as purification, which results in good-quality shared entanglement. This shared entanglement can then be used to send qubits over distances using teleportation. The quantum internet could be designed either to transmit qubits via one-way error correction or to provide good-quality entanglement as the base service. These services are informationally equivalent, but designing a network to do one or the other will require very different physical-layer designs, as well as the design of quantum network protocols. The benefits and resource trade-offs are unknown. Last, quantum communication is not possible without the assistance of classical communications. An open problem is understanding the optimal protocols needed to send both classical and quantum data over the same network at highest possible rates, Guha noted.
Photonic Quantum Computing
One of the biggest applications of a future quantum communications network is to enable distant quantum computers to talk to one another. That task may require quantum transduction between traveling qubits (photons) and matter qubits (e.g., superconducting qubits). It may be possible to bypass that requirement if the quantum computers themselves are using photons as qubits. Toward this end, Guha said, several researchers in Arizona, and those at two start-up companies, are developing all-photonic quantum computers using squeezed states of light, linear optics, and photon detectors. They are also researching scalable ways to build various quantum and classical integrated photonic devices with extremely
high fidelities necessary for quantum processing, and subsystems such as quantum repeaters and novel optical communications receivers.
A truly scalable quantum communications network is still a long way off, Guha asserted. A quantum communications network cannot work with classical amplifiers as quantum repeaters.14 Quantum versions of amplifiers have not yet been built. “Quantum repeater” is the term used to refer to a quantum processor that serves as an amplifier, switch, and router for quantum communications. It will be an essential component of the quantum internet, although there remain many unanswered questions that must be addressed in order to develop one. Two of the basic building blocks of a quantum repeater will be a quantum memory and a Bell State Measurement (BSM) that can convert 2 qubits held across two individual channel segments to a long-distance entanglement across the entire length of the two segments. More advanced repeaters could employ more complex quantum logic, such as 3-qubit joint measurements. However, communication rates will still be limited by the loss of the lossiest link in the channel. Even with noise and lossy devices to build a quantum repeater, it will be possible to attain a better end-to-end quantum communication rate than direct transmission, Guha posited.
Current experiments are moving toward building various designs of quantum repeaters. Each design involves some form of quantum memory, quantum and classical processing, and quantum and classical connections to neighboring nodes. There are many contenders for the qubit technology for the memory and processor, transduction, and choice of error correcting codes to build a quantum repeater. Instead of a matter-qubit-based quantum memory, several researchers are experimenting with a new concept called optical cluster states that could be used in lieu of a quantum memory to build a repeater, Guha said.
There are many open questions leading to the design of the future quantum internet. These include not only designing and building a repeater but also inherently networking problems such as what the optimal ways to design the quantum network are—that is, where to deploy the repeater nodes. There is also an entire open area of developing the quantum networking protocol stack, which will include
14 R. Namiki, O. Gittsovich, S. Guha, and N. Lütkenhaus, 2014, Gaussian-only regenerative stations cannot act as quantum repeaters, Physical Review A 90:062316, https://doi.org/10.1103/PhysRevA.90.062316.
switching, routing, and scheduling protocols to efficiently manage multiple end-to-end quantum communication flows through the quantum internet.
Broadly speaking, Guha said the development of quantum networking is in its early stages, but it is garnering rapid interest. Distributed quantum computing as a subject is also in its infancy. Developing a scalable quantum network and applications will require a highly interdisciplinary endeavor, involving research in physics of quantum materials, quantum integrated devices and photonics, quantum error correction, algorithms, and quantum computer networking. There are many eventual applications, including secure multiparty computations, distributed sensors enhanced by entanglement, and secure access to quantum computers on the cloud.
In response to a question about photonic quantum computing, Guha clarified that large continuous variable optical cluster states have been demonstrated, but they do not have the full power of quantum processing yet, but the technology is advancing fast. Continuous variable optical cluster states along with photon number resolving detectors will enable universal quantum computing, but more work is necessary to design the architectures. Researchers are also looking into discrete-variable photonic clusters. Although they are harder to generate, even 3-photon discrete-variable cluster sources will enable a highly efficient architecture for universal all-photonic quantum computing.
Nitin Samarth, Pennsylvania State University
Samarth, a condensed materials experimentalist, discussed the development of quantum materials, opportunities for their application in next-generation quantum technologies, and other issues in the quantum ecosystem.
First-Generation Quantum Materials and Technologies
Quantum materials are hard to define, but broadly speaking, they are solids with exotic physical properties due to the quantum mechanical properties of their constituent electrons. They have great scientific and technological potential, and they include key concepts that cannot be explained through the language of classical physics, such as quantum confinement, quantum fluctuations, and quantum entanglement.15
15 Office of Science, U.S. Department of Energy, 2016, Basic Research Needs Workshop on Quantum Materials for Energy Relevant Technology, https://science.osti.gov/-/media/bes/pdf/reports/2016/BRNQM_rpt_Final_12-09-2016.pdf?la=en&hash=E7760711641883FFC9F110D70385937D6A31C64F.
Quantum materials have had a large impact on first-generation quantum technologies—for instance, maximizing electron mobility through wave function engineering in heterostructure materials with low defect density. There is still much that is unknown, however. The rich landscape of fundamental physics, quantum materials, and current applications has led to intriguing concepts and great potential for quantum technology.
Most first-generation quantum technologies rely on quantum materials that exploit the manipulation of the amplitude of quantum wave functions. For example, semiconductor heterostructures in cell phones and magnetic multilayers for memory have been around for decades, while light-emitting semiconductor nanostructures are now used in solid state displays, Samarth noted.
Next-Generation Quantum Materials and Technologies
The next generation of quantum materials and technologies seeks to exploit all aspects of the quantum mechanical wave function, Samarth said, such as amplitude, phase, inherently nonlocal character, and entanglement. This is necessary for building quantum architectures and networks that exchange quantum information coherently between nodes and channels. This quantum transduction and quantum transfer process was first demonstrated using photons that convey quantum information between quantum states in systems such as cold atoms and trapped ions. Photon-based entanglement has also now been demonstrated in solid-state quantum information platforms such as superconducting qubits, quantum dot qubits, and defect spin qubits. While photons are certainly the most attractive vehicle for coherently transporting quantum information across long distances, Samarth said, there are opportunities for thinking about other types of quantum channels that could be exploited for communicating quantum information over shorter distances in chip-based architectures.16
According to Samarth, two key questions that materials scientists need to ask are: What materials—bulk crystals, thin films, or something else—will these new visions and applications in quantum engineering need? What would it take for entanglement to result from a nonphotonic mode even at short chip-scale distances?
To work on quantum materials, the materials themselves must first be synthesized, and different approaches may yield new useful materials. Bulk crystal growth in the United States continues to provide extremely high-quality quantum materials for the community. Leading bulk crystal growth laboratories include the National Science Foundation (NSF) Materials Innovation Platform (MIP) sites at
16 D.D. Awschalom, L.C. Bassett, A.S. Dzurak, E.L. Hu, and J.R. Petta, 2013, Quantum spintronics: Engineering and manipulating atom-like spins in semiconductors, Science 339(6124):1174-1179, https://doi.org/10.1126/science.1231364.
Pennsylvania State University, Johns Hopkins University, and Cornell University. These MIPs have also provided an opportunity to advance the state of the art in thin-film synthesis using molecular beam epitaxy (MBE) coupled with in vacuo characterization via advanced characterization tools. MBE is a particularly appealing approach to synthesizing quantum materials, Samarth said. For example, after decades of experience, it is now possible to grow exceptionally pure, near-perfect materials, such as gallium arsenide–based heterostructures, the quintessential topological quantum material. Other semiconductor heterostructures such as Si/SiGe have also been exploited for developing quantum dot spin qubits and even used for 2-bit quantum processors, Samarth noted.17 While perfect semiconductor materials are very useful for developing qubit platforms, even defective materials have proven to be important. Examples include nitrogen-vacancy (NV) centers in diamond and divacancy defects in SiC, both of which serve as single-spin qubits with long coherence times. NV center single-spin qubits have been used to test applications such as long-range entanglement and Bell’s inequality violations.
There have also been several early investigations of superconductor–semiconductor quantum materials for studying an emergent phenomenon in condensed matter known as the Majorana fermion (or Majorana zero mode). These studies are theoretically motivated by the application of Majorana modes for fault-tolerant topological quantum computing. While the theory is well established, much more work is needed to definitively prove the very existence of the Majorana mode, let alone its implementation as a topological qubit, Samarth said.
Although present-day quantum materials display fascinating quantum phenomena, exploiting these materials in quantum networks is still challenging. Defect qubits are clearly valuable in the context of photon-based networks, but a key question, Samarth said, is how to go beyond photons. Researchers are now beginning to explore hybrid materials, such as those combining topology and superconductivity, those coupling NV centers in nanodiamonds with magnons (spin waves) in ferrimagnets,18 and those using MBE growth to interface magnetic order with topological states. These approaches could provide a means to exchange quantum information between a localized qubit and a flying qubit in an on-chip channel. Novel synthesis of quantum materials might provide attractive routes toward
17 L.M.K. Vandersypen, H. Bluhm, J.S. Clarke, A.S. Dzurak, R. Ishihara, A. Morello, D.J. Reilly, L.R. Schreiber, and M. Veldhorst, 2017, Interfacing spin qubits in quantum dots and donors—hot, dense, and coherent, npj Quantum Information 3(34), https://doi.org/10.1038/s41534-017-0038-y.
18 D.M. Toyli, C.D. Weis, G.D. Fuchs, T. Schenkel, and D.D. Awschalom, 2010, Chip-scale nanofabrication of single spins and spin arrays in diamond, Nano Letters 10(8):3168-3172, https://pubs.acs.org/doi/10.1021/nl102066q.
Majorana modes by coupling topological states with superconductivity, Samarth posited.
Before concluding, Samarth noted that the quantum materials field is lacking students and postdoctoral researchers with adequate training in materials synthesis, yet materials knowledge is critical to creating a skilled quantum science and technology workforce. Materials synthesis and characterization will lead to the development of new quantum technologies, he posited, arguing that until it is clear what that future technology will require, all materials research areas should remain open. New approaches with new materials may one day incorporate coherence and entanglement, solving old problems while also creating new ones.
Asked if it is possible to exploit the long history of studies of defects in conventional materials, Samarth answered in the affirmative, noting that researchers are indeed trying to model defects in a variety of materials to identify which ones would be most interesting to explore from the viewpoint of quantum technologies. Although placing defects in solid-state platforms in a deterministic manner poses a significant challenge, some advances have already been made (e.g., at Argonne National Laboratory and the University of Chicago). Asked if it may be better to instead engineer our way around stochastically located defects, Samarth also agreed that this would be a route worth considering.