Homeland Security Advisory Council - Dhs

which can implement QKD with any other node on the network. These networks must be one-toone as QKD inherently cannot be done in such a way to guarantee that the key generated betweentwo nodes will be the same as guaranteed between any other two nodes. It is not possible for onenode to share the same key with multiple other nodes.The second type of quantum network is a network in which the various nodes share entangledphotons. As described above, the entanglement can serve as a resource enabling the nodes tocommunicate without fear of an eavesdropper. Sharing entanglement over long distances is not aneasy task due to absorption in fiber. One may choose to employ a quantum repeater to boost thedistance over which entanglement can be shared. A quantum repeater, unlike classical repeaters,cannot amplify. Instead, quantum repeaters measure entangled photons in such a way as to allowentanglement to be shared over longer distances.Quantum SensorsQuantum phenomena have the potential to enhance several quantum sensing modalities includingmagnetometry, electrometry, gravimetery, and associated techniques such as accelerometers andgyroscopes. Three platforms of interest for quantum sensors are atoms, artificial atoms, and light. Ashort introduction to each is included below.Atomic sensors can operate in two different ways: as interferometers, using beams of atoms sentthrough atomic interferometers where, due to their extremely short wavelength they are moresensitive than typical light-based interferometers, and as sensors of magnetic and electric fields.Atomic interferometers can be built such that they are sensitive to different types of accelerationsand the presence of mass. For both, one path of the interferometer will be more affected than theother, causing a path-length difference. Hence, they can be used as gravimeters, gravitygradiometers, accelerometers, and gyroscopes. All have the potential to improve upon theperformance of classical sensors in both sensitivity and long-term accuracy. They are also expectedto improve upon alternative quantum sensors (superconducting quantum interference device or“SQUIDs”) in both Size, Weight, and Power (SWaP) and long-term accuracy. 6 Potentialapplications for atom-interferometer based sensors include accurate position, navigation, andtiming (PNT) for navigation in GPS-denied environments, tunnel detection, detection of WMD,and detection of anomalously heavy objects (e.g., loaded trucks, shipping containers).Atomic magnetometers detect, measure, and assess magnetic fields based on the precession rate ofthe atom. These systems have already demonstrated sensitivity below a femto-Tesla in thelaboratory and have a very low SWAP C compared with today’s best technologies. In addition,these systems can be placed close to a potential source and thus could revolutionize magneticresonance imaging (MRI).Atomic electrometers are atoms with a highly excited outer electron allowing them to behave ashighly sensitive electric dipoles. These systems can achieve much higher sensitivity than standard6SQUIDs are highly sensitive magnetometers used to measure tiny magnetic fields, key for many scientific uses likebiology and magnetic property measurement systems. They have both civilian and military uses.13 P a g e

antennas and they have the potential to sense carrier frequencies from DC through 10 THz(10 1012 Hz) without changing the physical platform. They do not follow the standard Chu limit(i.e., their bandwidth is not limited by the size of the sensors), and they inherently reject out-ofband noise. Unlike standard antennas, they do not absorb the field that they are detecting. Despitetheir current limitations in bandwidth and dynamic range, atomic electrometers are uniquelypositioned to enable novel electric field sensing modalities including communications in theuntapped THz band for terabits/sec data rates, a portable calibration standard for THz frequencies(which does not yet exist), and sub-wavelength field mapping and imaging over a broad spectralrange. These “atom radios” will enable high-rate free-space communications, which could be usedas secure nodes for PNT or for high-vision data delivery for telemedicine.Defects in solid-state systems, a mis-placed atom in a crystal of other atoms, can act as artificialatoms and be used as sensors. An example is nitrogen-vacancy (NV) centers in which a nitrogenatom displaces two carbon atoms in a diamond lattice. The advantage of these sensors is their highsensitivity, which is at least partly due to their extremely low SWAP, allowing them to be placedextremely close to the sample to be tested. In addition, because these systems are in a solid-statelattice, they have the ability to be placed in biological and other environments where other sensorswould not be tolerated.When photons (particles of light) are entangled, they can be used as probes for increased sensingcapability. The first example of this was quantum lithography, in which entangled light is used tomake smaller lithographic lines than typical photons. Quantum illumination is a technology that canpotentially use entangled photons to increase signal-to-noise ratio for radar and lidar by answeringthe question of whether or not there is a target at a pre-specified distance (in a given direction) fromthe sensor. One of the pair of photons is sent towards the target, while the other is stored until thefirst returns. They are then jointly measured. This technology may be useful for situations in whicha warning receiver—used by an adversary to detect radio emissions of radar systems—would limitthe ability to sense a target.Quantum illumination can enhance the sensitivity of both lidar and radar systems by improvingtheir ability to detect faint objects against noisy backgrounds. While there have been someproposals to use this technique for close-range biomedical imaging, the most commonly proposedapplication is quantum radar, which improves radar systems’ ability to detect faint objects at longrange.In principle, quantum radar (or any other form of quantum illumination) can achieve a signal-tonoise ratio that is 6 dB higher than the best standard radar system. 7 But useful quantum radarsystems have proven very difficult to engineer in practice. There are several fundamental obstaclesto the useful deployment of quantum illumination and radar: for example, the transceivers must becooled down to extremely low temperatures, and the fact that the target distance must be known inadvance limits the technique’s utility for detecting objects at long distances.7Si-Hui Tan et al., “Quantum Illumination with Gaussian States,” Physical Review Letters 101, 253610 (2008).14 P a g e

1. Assessment of the current state and perceived future advancements over the next 3-10years that could pose a threat to the homeland security of the United States.Various governments, large multi-national corporations and innumerable startups are engaged in theresearch and development of quantum technologies. In what has been deemed the “Quantum Race,”countries worldwide more than tripled their investment in quantum computing and software between2012 and 2019. 8Today, the United States and China are the greatest investors, with other countries following closebehind. As of 2018, the total U.S. federal government investment in quantum technology R&D wasestimated to be 200-250 million per year. 9 U.S. investment is currently increasing significantly: in2018 the President signed into law the National Quantum Initiative Act, which authorized 1.275billion over five years for the Department of Energy (DOE), the National Science Foundation(NSF), and the National Institute for Standards and Technology (NIST) to invest in R&D inquantum information science.In 2016, President Xi Jinping of China established a national strategy for the country to becometechnologically self-reliant. The country has stepped up its research and investment in quantumtechnologies significantly since then, hoping to surpass the United States as the leader intechnology. 10 China’s government was estimated to be investing 244 million in quantum R&D peryear as of 2018. 11 As of January 2019, the US Senate’s Worldwide Threat Assessment report statedthat the United States’ lead in technology had been significantly eroded, mainly by China. 12 TheU.S.-China Economic and Security Review Commission has concluded that “China has closed thetechnological gap with the United States in quantum information science—a sector the UnitedStates has long dominated.” 13The UK and the European Union, particularly Germany, France, and the Netherlands, havecommitted to invest heavily in quantum technology over the next ten years. The UK, for example,lined up 193 million worth of investments and commitments from industry players in 2019,bringing total funding to over 440 million. 14 The European Union has pledged 1.1 billion inPaul Smith Goodson, “Quantum USA Vs. Quantum China: The World's Most Important Technology Race,” Forbes,October 10, 2019, ortant-technology-race/?sh 1ae9fdd472de.9Congressional Research Service, “Federal Quantum Information Science: An Overview,” July 2, dson.11Congressional Research Service, 2018.12Daniel R. Coats, Director of National Intelligence, “Statement for the Record: Worldwide Threat Assessment of theUS Intelligence Community,” Senate Select Committee on Intelligence, January 29, -ATA-SFR---SSCI.pdf.13U.S.-China Economic and Security Review Commission, 2017 Report to Congress of the U.S.-China Economic andSecurity Review Commission, Washington, D.C.: U.S. Government Publishing Office, 2017, pg. report-congress14Smriti Srivastava, “Top 10 Countries Leading in Quantum Computing Technology,” Analytic Insight, December 14,2019, leading-quantum-computing-technology/.815 P a g e

government investment over 10 years. 15 Japan and South Korea are making smaller but stillsignificant investments of tens of millions of dollars per year. 16In recent years, a robust private sector in applied quantum information science has developed, withcompanies in U.S., Canada, the E.U., and Australia making hundreds of millions of dollars inprivate-sector investment. There appear to be many fewer private-sector companies in China andJapan investing in quantum information science R&D. 17Below we provide some detail of the current state of each quantum technology and possible nearterm advancement.1.1 Quantum ComputersTo date, there are no quantum computers capable of large-scale implementations of either of theabove-mentioned algorithms. However, several large corporations and startups have made significantinvestments in the area, resulting in small scale quantum computers on the order of 50 qubits (thequantum parallel to a classical bit). IBM, Intel, and Google have all announced the construction ofquantum computers on the order of 50 superconducting qubits, while Rigetti and Alibaba are around20 superconducting qubits. The startup IonQ has recently announced a system with 79 ion qubits. Inaddition, IBM, Alibaba and a startup called Rigetti allow for cloud access to their small-scalequantum computers. While this may not sound like much, it is generally agreed that a quantumcomputer with on the order of 100 qubits could perform certain algorithms that are not possible evenfor today’s supercomputers.While recent progress is encouraging, there is still a long way to go before a quantum computerwould be capable of breaking RSA. Current systems with a few tens of qubits can perform, at most,a few tens of basic logic operations (known as logic “gates”). By way of comparison, for a quantumcomputer to break elliptic curve cryptography (with a security factor of 300) would require about2700 qubits and 1.8X1011 gates, and that is assuming gates are implemented perfectly. A concretemetric properly characterizing the readiness of a quantum computer is still lacking.While almost all experts agree that there will be highly capable quantum computers at some point inthe future, estimates as to when a quantum computer capable of breaking public key encryption willcome on line ranges from 10 to 50 years. The National Academy of Sciences issued an expertconsensus that quantum computers will not threaten encryption over the next 10 years. 18 However,even in the near term of 3 – 10 years, where the number of qubits is limited and the implementationsare far from optimal, there may be some problems of interest that could be tackled.Congressional Research Service, 2018.Jason Palmer, “Here, There, and Everywhere,” The Economist, beth Gibney, “Quantum Gold Rush: The Private Funding Pouring into Quantum Start- Ups,” Nature, Vol. 574,2019, pp. 22–24.18National Academies of Sciences, Engineering, and Medicine. 2019. Quantum Computing: Progress and Prospects.Washington, DC: The National Academies Press. https://doi.org/10.17226/25196.151616 P a g e

Quantum Machine LearningQuantum computers can theoretically perform various machine learning techniques more quickly,more accurately, and with less training than classical computers. These include supervised learningtechniques such as neural networks and support vector machines for image classification, andunsupervised learning problems such as clustering. Quantum Machine Learning (QML) is currentlyone of the fastest growing research areas, and a new journal dedicated to the field has just comeonline (a Springer journal called Quantum Machine Intelligence). Experimental tests of QMLtechniques have already been performed on the Rigetti 19-qubit system and on the D-Wave systemdiscussed below. Finding the optimal solution is not necessary for machine learning; instead the goalis to find a good (better than classical) answer in a short period of time, and many feel that errorswhich would derail a quantum computer performing a typical algorithm may not cause catastrophicfailure to QML. Instead, errors would simply cause the answer to be less optimal. Whether thisfeeling is accurate is still open to question and requires additional research.Quantum SimulationsSimulating quantum systems on classical computers is inefficient and extremely difficult. RichardFeynman was the first to note that the proper way to simulate a quantum system is to use anotherquantum system that we can control. Used in this way a quantum computer could revolutionizecomputational material science and perhaps provide needed insight into high-temperaturesuperconducting phenomenon and nitrogenases. Like QML, there are those who feel that quantumsimulations may be possible in a nearer timeframe than other quantum algorithms. This is becausewe do not expect systems to be simulated to be in states as complex as the qubits of a quantumcomputer during a typical algorithm. Again, whether this is accurate requires additional research.Quantum AnnealersQuantum annealers are systems that exploit quantum phenomena to solve a more limited class ofproblems than a universal quantum computer. The quantum annealer machines produced by thecompany D-Wave have been demonstrated to solve optimization problems ranging from trafficpatterns to materials analysis, and quantum machine learning to scheduling and navigation problems.While it has been shown that for certain problems the D-Wave system can outperform specificoptimization routines, it has yet to be shown that it can outperform all classical heuristics.Nevertheless, it is possible that in the near-term these systems will prove to be both more efficientthan many classical systems and more cost-effective than large-scale quantum computers.A key distinction lies in different implementations of quantu

Introduction Quantum information science (QIS) is the field dedicated to exploiting quantum phenomena for the . QIS can be divided into three areas: quantum computing, quantum communications, and quantum sensing. Quantum computers, if fully developed, could break all currently used public key