Behold, the Quantum Network Cometh

I read yesterday that Cisco has announced its theoretical and practical designs on quantum networking. It makes sense. Cisco was and is the industry behemoth of what we’ll call classical networking, but quantum networking — with all due respect to the networking optimizations required for artificial intelligence — is likely to be the next big thing in networking.

Perhaps quantum networking will be very big indeed, but its arrival isn’t imminent. There’s time for us to learn about quantum networking before it gets here. That said, Cisco and others see no harm (and considerable potential benefit) in establishing real or perceived thought leadership in an emerging area that can only grow commercially.

Okay, let’s grapple with the beast. Alas, quantum networking: my head hurts just thinking about it.

Still, if one looks ahead to the next major chapter of networking — now that cloud-native technologies have surmounted the application and service layers of the traditional network stack — quantum networking seems an inevitability. When it arrives, quantum networking will introduce infrastructure, mechanisms, protocols, and solutions that are distinctly different from those belonging to classical networks.

Where there are problems, there are opportunities. Networking problems abound in quantum computing, and classical approaches will not suffice — not on their own, anyway.

I’m neither a quantum physicist nor a quantum engineer. I’m more like a disheveled Lieutenant Columbo wandering around the quantum laboratory and importuning the experts and specialists with annoying questions. Occasionally, as I consider their replies, a brief spark of cognition passes through my cerebral cortex, like the flashing lights of a pinball machine or the goal light at a hockey game. Those lights flash, but can they stay on? Well, let me see whether I can provide enough juice to keep them burning, allowing me to learn and maybe even impart something potentially useful about the subject at hand.

Just One More Thing . . .

I’ve included the foregoing as a public-service announcement. I caution that what follows should be taken not as an authoritative professorial lecture, but more like the tentative advances of a curious newcomer. Like Columbo, I’m blundering and shambling toward new insights, but unlike the vintage television sleuth, I’m not about to solve the case conclusively or experience an arresting epiphany before today’s show concludes. If I’m fortunate, I’ll find myself a little further along the road of discovery.

Anyway, here’s my admittedly rudimentary understanding of quantum networking, including my assessment of why it will be markedly different from classical data networking. This is a work in progress, of course. And so am I.

Quantum networking introduces fundamentally different connectivity requirements, above and beyond those addressed by classical networking. That’s because quantum networking must accommodate the peculiar characteristics and properties of quantum information.

What follows is an enumeration of the areas where quantum networking must, by necessity, go beyond the capabilities of classical data networking.

First up is what’s called quantum state preservation. Why is that a requirement? Unlike bits in classical networking, quantum bits (qubits) exist in fragile superposition states, and it is the quantum network’s job to maintain quantum coherence during qubit transmission. By contrast, classical networks need only preserve the integrity of packetized discrete 0s and 1s.

Another unique challenge in quantum networking is entanglement distribution. In the quantum world, processing and network infrastructure must create and distribute entangled qubits. Entanglement enables and supports protocols such as quantum teleportation and dense coding. In classical networking, as you might have deduced, no analogous requirement exists.

Moreover, pursuant to what is called the no-cloning theorem, quantum information cannot be copied. In classical networking, of course, data packets are copied, often extensively, for purposes such as routing and error correction. In quantum networking, fundamentally new approaches are required to address error correction. Routing also requires new approaches — more on which later.

Rise of the Quantum Network

These quantum-networking requirements set the stage for new types of physical network infrastructure, most of which complements or augments traditional network infrastructure.

Among this new infrastructure we find quantum repeaters. Whereas classical repeaters amplify signals across distances, quantum repeaters must preserve entanglement across distances. To do so, quantum repeaters use quantum memory and quantum-swapping functions.

As for quantum memory, it gives rise to quantum’s answer to storage networking. Classical networks, as we know, support various permutations of digital storage. Quantum networks, however, require specialized quantum memory that stores qubits while maintaining coherence. Quantum memory requirements are addressed by methods such as trapped atoms, ions, or solid-state systems.

Specialized transmission media will also be needed. In classical networks, we have the likes of copper, fiber optics, and wireless media. Those will be complemented or augmented by optical fiber that will be capable of meeting the exacting loss requirements of photonic qubits. In addition, high-performance satellite links are likely to be used for long-distance quantum communications.

As you’ve probably gathered, we’re straying far from the familiar confines of Ethernet, Fibre Channel, InfiniBand, IP, standard approaches to routing, and TCP/UDP at the transport layer. This is not the OSI model, though we will subsequently consider the form a quantum-networking stack might take.

Given its unique mandate, quantum networking requires its own protocols.

For example, classical protocols for key exchange are not suited to quantum requirements, which call upon quantum properties to establish secure encryption keys. Quantum calls for information-theoretic security rather than computational security.

Let’s now get back to routing, a particularly engrossing quantum challenge. Quantum information would be obliterated by conventional packet-inspection techniques. Hence, those are not available. In quantum networks, routes are often predetermined or use other methods that do not require inspection.

Indeed, quantum routing is not amenable to classical forms of network management or monitoring. The constraints of quantum routing introduce particular challenges. As we know, conventional routers use packet headers to make routing decisions. If quantum information is “read,” it collapses or otherwise dissipates. This was discussed briefly above. This means that quantum router nodes are proscribed from determining destinations by inspecting quantum information, enforcing a no-read requirement on routing protocols.

Another complication is presented by quantum’s no-cloning constraint. In broadcast and multicast scenarios, classical network infrastructure duplicates packets. A no-cloning constraint in quantum computing and networking prohibits duplication, meaning that each qubit must follow a single path through the network.

Keep Calm and Let Quantum Carry On

Quantum routing also insists upon treating entanglement as an intrinsic resource. I’d like to explain this routing problem to you in more detail, but I can’t. It’s beyond my limited understanding. Suffice it to say, no classical analog exists for quantum’s entanglement-resource management requirement.

The entanglement concept is, for me, baffling. For a long time, as this article published in New Scientist makes evident, I wasn’t alone in my discomposure. Much bigger brains than mine have struggled to comprehend the logical consistency of entanglement.

While scientists generally try to find sensible explanations for weird phenomena, quantum entanglement has them tied in knots.
This link between subatomic particles, in which they appear to instantly influence one another no matter how far apart, defies our understanding of space and time. It famously confounded Albert Einstein, who dubbed it “spooky action at a distance”. And it continues to be a source of mystery today. “These quantum correlations seem to appear somehow from outside space-time, in the sense that there is no story in space and time that explains them,” says Nicolas Gisin at the University of Geneva, Switzerland.
But the truth is that, as physicists have come to accept the mysterious nature of entanglement and are using it to develop new technologies, they are doubtful that it has anything left to tell us about how the universe works.

Well, that’s a relief. I should keep calm and carry on with entanglement.

In my next missive, to be transported across classical networks tomorrow, I’ll discuss the architectural construct of a quantum network stack. I will also attempt to explain, if space and time allow, why conventional content distribution networks (CDNs) will be supplanted or substantially enhanced by new types of infrastructure designed to facilitate efficient distribution of quantum information.