22.10.2019by Bas Dirkse
Quantum internet: at the verge of an emerging technology
by Bas Dirkse
The internet as we know it today has become an integral part of our lives. We use this piece of technology on a daily, hourly, almost continuous basis. We use it at work, to relax, to socialize, to fact check our friends during an argument and even to control the thermostat at our homes. We can definitely claim that, despite the internet bubble in the 90’s, the technology has far outperformed its expectations of the early 60’s and 70’s in the societal and financial benefits it provided. Will the same bright future be reserved for the quantum internet? Will it deliver the same amount of societal and financial benefits to the world as its classical sibling has done?
If a quantum internet is something you have never heard about, don’t worry! You are not lagging behind or living under a rock. This technology is something that today only really exists on paper and on the drawing boards of scientists. Yet, I argue, it is now a great moment to start discussing its potential, as a lot of research effort is being made towards the realization of such a quantum internet. So first of all, what is this new potential piece of technology even about? What is a quantum internet?
What is a quantum internet?
This question can be addressed in many different ways and on many different levels of technical detail (see e.g. the blog post by Matteo,  or  for different definitions). There is no agreed upon definition of a quantum internet, but I will contribute an attempt. For the purpose of this article, I will take an operational approach. To me, a (general purpose) quantum internet is a network of physically separated quantum computers that can deliver an entangled pair of quantum bits (qubits) on demand to any pair of connected end-node quantum computers (each end-node receives one qubit from the pair) . These end-node quantum computers need not be very powerful, but many interesting applications require them to at least have some capabilities (e.g. storing a few qubits and operating on them ).
In a quantum internet, entanglement is the crucial resource (and the defining feature). If the two shared qubits were not entangled, this network would be quite useless. Two qubits are entangled if their quantum state can only be described jointly. In contrast, in a non-entangled pair of qubits, the state of each qubit can be described independently of the other (the concept of entanglement is very nicely explained in a previous blog post by Matteo). This shared entanglement does not come for free. In the deep, inner layers of the working of a quantum internet, a lot of coordination and communication is needed to produce entanglement on demand to end-users (as explained in the post by Axel).
So what is this quantum internet for?
Let me start answering this question by saying what it is not for. The purpose of a quantum internet is not to replace, improve or outperform the current ‘classical’ internet entirely. To the contrary, a quantum internet will operate in great synergy with the classical internet. Regular ‘classical’ communication and applications will continue to run on the internet as we are used to. The quantum internet will only be utilized for specialized applications and subroutines where it outperforms the internet in some way. This is similar to how a quantum computer will likely never replace a classical computer entirely – it will outperform a classical computer only for very specific problems. Moreover, many specialized quantum internet applications also require communication of classical bits over the standard internet.
Ok, so a quantum internet will be a specialized type of network that can outperform the internet on particular specialized tasks. Perhaps it is at this point unclear why such tasks should even exist:
Well, one answer is that we already know of several examples where the quantum internet achieves things that are impossible with only the ‘classical’ internet. The most well-known example is quantum key distribution (more on this later). But first, let me explain why in principle a quantum internet can have an advantage over a classical internet.
An advantage of a quantum internet over its classical counterpart can be obtained by exploiting its unique capability: sharing entanglement between remote end-users. Indeed, many applications exploit the exotic properties of this shared entanglement. There are two fundamental properties of entanglement on which most specialized quantum internet applications rely. First, shared entanglement is inherently private. If two qubits are maximally entangled, then a third must be completely separated from the entangled pair. The entanglement is private to the two end-users holding it. This is exploited in many cryptographic applications. Second, shared entanglement allows for maximal coordination. If I measure one qubit and I get a particular outcome, then the same measurement on the other qubit will always result in the same outcome, even if this outcome was not determined beforehand. Many applications known today exploit one or both of these fundamental properties to some degree.
What applications do we know of already?
Up until know, I have hinted at the existence of several applications in which a quantum internet outperforms the current, classical internet we have today. The first and most well-known example of this is quantum key distribution (QKD) . QKD allows end-users to share a (classical) secret key of which the security is guaranteed by the laws of (quantum) physics only. This is impossible using the internet, where some type of assumption is always necessary (e.g. RSA encryption relies on the assumption that factoring prime numbers is a hard problem for computers – an assumption that breaks down when quantum computers become available). The strength of the security then depends on the strength of the assumption. Shared secret keys have a variety of applications, for example encryption of private communication. Hence with QKD, long term secrecy of communicated data can be ensured by the laws of quantum mechanics.
There are many other applications known today in the domain of cryptography, which are less well-known than QKD. For example, quantum protocols are known for secure identification. The task of identification on a network is ubiquitous: every time you log into your bank account, you identify yourself by entering a password that is only known to you. Quantum protocols can enhance the security for this task too. Currently, the security of authentication is based on computational assumptions, which may be retroactively broken. Moreover, additional work must be put in place to prevent man-in-the-middle attacks. With quantum secure identification protocols, the security only relies on the physical assumption that storing many qubits for a long time is difficult. Moreover, after the identification is completed, this assumption need no longer hold in the future (in contrast to classical encryption) and the quantum protocol immediately protects also against a man-in-the-middle attack. Thus, logging into remote applications can be made much more secure with quantum communication available.
A third interesting application of a quantum internet is delegated quantum computing. In this task, an end-user with only a (very) limited quantum computer can delegate a complicated quantum computation to a remote, large quantum computer. Of course, this requires the development of large-scale quantum computers to delegate computations to. The client with the small computer can send input states and instructions to the server over the quantum internet. Variants exist in which the input (quantum) data and the computation itself can be encrypted (no person including the server can learn the input and computation) and verified (the client can verify that the server honestly performed the computation requested) .
Finally, there are very interesting known applications of quantum networks for sensing and metrology tasks, in which the goal is to measure a physical quantity very accurately. An example of this is improved clock synchronization of remote atomic clocks by consuming shared entanglement . This would allow atomic clocks all over the world to keep time more precise than is possible without shared entanglement. This is relevant for many applications where precise timing information is important. For example, more precise timing information would directly allow for more precise GPS systems. This is because GPS determines your position by computing your distance to various satellites, which determined by the time it takes to for a message to travel from the satellite to your GPS receiver. More accurate clocks will allow more accurate measurement of the time-of-travel of these messages, hence allowing more accurate distance calculation.
A second example of improved metrology is extending the baseline of telescopes using shared entanglement as a resource . Concretely this would mean that telescopes can become much better with entanglement, allowing much better resolution of optical images taken from outer space. The scientific impact of this would be significant. In this domain of application, the maximal coordination property of shared entanglement is exploited.
How will these applications impact our society?
“All of these applications sound promising and interesting”, you say, “but when would I want to establish secret keys or delegate a quantum computation?” In other words, how will we as a society benefit from these applications and when? These are important questions. The future potential of the quantum internet as an emerging technology heavily depends on the demand there is for it. Demand for this technology is generated from industry, society and governments if there are real-world problems that a quantum internet can solve (more efficiently than other methods). So during the development of the quantum internet, it is crucial is to keep exploring and identifying where known (and new) applications of a quantum internet can solve real-word business or societal problems. This not only continues to justify investments (both from industry and from society) into the development of a quantum internet, but can also guide the development into a certain direction.
To my surprise, these societal impact questions are addressed in much less detail for the quantum internet than for quantum computing, where, industry, government and academia come together to think about the future impact of this technology (see for example this study by BCG estimating the future market value of quantum computing). The likely reason for this is that quantum computing devices (however noisy and small-scale they may be), are available today to a broad audience (e.g. online in the cloud). In contrast, the quantum internet is just not quite there yet, with the first true general purpose quantum internet being on the verge of existence. But an insightful contribution to societal and commercial impact of a quantum internet has already been published by TU Delft in the magazine Quantum Internet (June 2019). This 60 page magazine gives a very accessible overview of the basics of quantum networks, currently known applications and their impact on society and governance. Interestingly, they also examine potential risks to society imposed by the advent of the quantum internet.
Specifically for QKD, impact and market studies are a bit more common (see e.g. this commercial report on the market development of QKD the upcoming decade – April 2019). This is likely because QKD is already available as a commercial technology. Commercial QKD technology is dedicated specifically and exclusively to the task of QKD, and is not performed over a general quantum internet as I discuss here. This allowed the technology to develop more rapidly. The estimated market value of QKD is $85 million in 2019 and is expected to grow to $980 million in 2024. So QKD is already an application that adds real-world value to companies and governments for which encryption is of utmost importance. For example, QKD is being used since 2007 to protect the election results in the state of Geneva, Switzerland or by a Swiss bank to securely back up data to an off-site data recovery center.
It seems to be common agreement that enhanced quantum cryptography is the first major area of application where the quantum internet will contribute to society. This has several reasons. First, quantum cryptography (including QKD and secure identification discussed above) require relatively little quantum capabilities and can therefore be executed on early stage quantum networks. Second, privacy and cybersecurity are important public values. A lot of effort is put in ensuring privacy and security in various sorts of (communication) networks, including the internet. Enhancing the level of security with quantum communication can therefore potentially be extremely valuable to society. Third, the advance of quantum technology itself poses potential threads to current practices in cryptography and therefore new solutions are necessary to protect our data in the future. Although quantum cryptography is not the only option (a lot of effort is also put into post-quantum cryptography in which security is guaranteed by assuming certain problems are hard even for powerful quantum computers), it does offer the strongest form of security especially against quantum attacks.
Beyond cryptography it is less known what the societal impact of quantum networks will be. However, as I mentioned before, areas of applications are known where quantum networks can contribute. As of now, it is unclear how far these applications will develop and when they will provide real-word value to society. So, let’s look a bit further in the future and imagine what a full-fletched quantum internet could contribute beyond everlasting, unhackable security.
Will the quantum internet revolutionize society the same way the classical internet did?
Of course, this is a hard question to answer and no one will know for sure. And the internet was a major revolution to society, so matching this is a tall order. But there are at least positive signs that the quantum internet can have a large impact to our future society. There is a lot of interest from society and industry, which is expressed in collaboration and investments into the research developing the quantum internet. For example, the European Union is investing the Quantum Internet Alliance, a large collaboration with many European research and industrial partners, with a long-term goal to develop a global quantum internet. Part of this project is also identification of real-world use cases for a quantum internet.
Moreover, there are already several promising potential applications in different areas. However, this is most likely just the tip of the iceberg. I suspect that the development of new applications will get an enormous boost once there is an actual (test-bed/small-scale) quantum internet available to prototype on. Interest from industry will also increase. This would lead to a much wider audience thinking about potential problems and use cases for quantum networks. Perhaps more companies will start issuing challenges and call to the community to solve their problems on a quantum internet (like this quantum internet hackaton). This would stimulate researchers from academia and industry to come up with new and relevant applications of a quantum internet.
But perhaps the biggest impact of the quantum internet on society will only be discovered after the technology has been developed and matured. This is not an unrealistic prediction; it has happened before with many different technologies (e.g. lasers). Actually, the classical internet is a striking example of this. In the early development of the internet (see ARPA-net), many of today’s use cases of the internet were inconceivable at the time. Of course, this is a bit speculative and there are different future scenario’s possible for this emerging technology.
The quantum internet could end up as (just?) a dedicated network that is used only by research and industry for very specific problems. But it might also be the case that decades from now all of us connect and communicate over a hybrid quantum/classical internet, with guaranteed security and enhanced performance every time we use the network to communicate, relax or socialize. Imagine a world where our phones, computers and wearables seamlessly connect and integrate with quantum technology to enables us with unparalleled new capabilities. The history of the classical internet over the past five decades tells us that this is at least a possibility. But only the future will tell in the end. In the meantime, I encourage the scientific community, industry and policy makers to continue exploring their potential needs for this revolutionary and promising new piece of technology.
Footnotes and References
 Maybe it seems weird that I don’t define a quantum internet as a network capable of communication qubits of information between any pair of network nodes (since the internet is a network that allows communication of bits between end-users). However, sharing entanglement is equivalent: if I can communicate qubits, I can locally entangle two qubits and send one to the other end to make entanglement and conversely, if I have an entangled pair of qubits, I can transmit an arbitrary qubit using quantum teleportation. Because sending a qubit over a noisy channel can cause you to lose precious information, the approach is to distribute and test the generation of known entangled states over a noisy channel until succeeded and `consume’ it to send the qubit without loss using quantum teleportation.
Bas is in the second year of his PhD in the theory group of Stephanie Wehner here at Qutech and the group of Michael Walter at QuSoft in Amsterdam. His research focuses on the applications of a quantum internet, identifying crucial building blocks and developing new application in the domain of distributed systems. After work, he enjoys to play baseball and board games.