Quantum Networks: Cisco Software Boosts Classical Tech

With the goal of creating a practical quantum computer, researchers are developing bigger and better quantum networks with capabilities that will complement and improve quantum computing. In other words, building a working quantum network capable of exchanging many qubits securely, over long distances, could be a useful end goal, completely independent of the quantum computer race.

With this in mind, Cisco launched a quantum network software system on September 25. The networking giant’s technology could help create more powerful quantum sensors, secure location verification, and quantum-enhanced imaging technology, to name just three of a range of emerging non-computing applications for quantum networks.

The team also has a hybrid goal in mind, says Ramana Kompella, vice president and head of research at Cisco in San Jose, Calif.: quantum networks capable of working with classical computers and conventional computer networks.

“This is a very fascinating area for us because until now, classical computing did not have access to a quantum network,” says Kompella. “But imagine if you had access to a quantum network, what could you actually offer in terms of new capabilities? »Kompella has a respond to his own question. “We can secure classical networks using quantum signals by detecting eavesdropping on long-distance optical fiber communications,” he explains.

How does quantum entanglement secure networks?

To do this, Kompella explains, the system relies on the fact that the quantum signals shared on their sensitive network are connected to each other via quantum entanglement. “We inject entangled photons into the optical fiber,” he explains. “And if the attacker tries to exploit the fiber, they end up disrupting the entanglement, which allows us to detect it.”

Kompella adds that entanglement traded over network distances has other classic computing applications in high-frequency trading and fintech, “and maybe you can also do ultra-precise time synchronization using entanglement-based networks,” he says.

Cisco’s quantum network system is built on a practical quantum network chip that the company introduced in May, which uses existing fiber optic lines, generates up to 200 million pairs of entangled photons per second and operates at standard telecommunications wavelengths.

But the new component recently introduced by Cisco is software. The compiler the company just launched allows a coder to write in IBM’s Python-based Qiskit quantum computing language. And the Cisco compiler takes care of the technical details of the network, like optimizing connections between quantum processors and fine-tuning error correction strategies.

“We hide the complexity of the physical layer,” says Reza Nejabati, head of quantum research at Cisco, “which allows algorithm developers to play with the number of processors and how the processors are connected together to optimize their algorithms.”

“The compiler takes that high-level goal, breaks it down, and then handles the networking side of the equation,” Kompella adds.

Hoi-Kwong Lo, a professor of electrical and computer engineering at the University of Toronto, says Cisco is championing an underappreciated part of the larger world of quantum technology.

“Investment is a key issue,” says Lo. “While billions in research funds are invested in quantum computing startups each year, investment in quantum network startups is lagging behind.”

According to Ronald Hanson, a professor of nanoscience at Delft University of Technology in the Netherlands, Cisco’s work represents a key next step. But this is just a next step.

“What Cisco is introducing today is not really the first of its kind,” Hanson says. “But the fact that Cisco is working on many of these different elements of the quantum network, combined with its expertise and strengths in classical networks, makes the progress interesting and will advance the quantum network industry as a whole.”

What will it take for quantum networks to scale?

According to Nejabati, the biggest current limitation of the Cisco system is the physical distance that a single photon can travel before being absorbed by the optical fiber itself.

“Our hardware and software technology allows us to… travel up to a hundred kilometers with a very high quality and efficient network,” explains Nejabati.

Lo says that physics – particularly a law called the “no-cloning theorem,” stating that individual quantum bits can never be perfectly replicated – makes large-scale quantum networks particularly difficult to achieve.

“The big challenge is building quantum repeaters,” says Lo. “Optical fibers cause losses and to overcome the distance limit we need quantum repeaters.”

Lo’s group, for example, studies the encoding of a qubit’s signal not on another individual photon, but rather on a group of entangled photons. IEEE Spectrum followed Lo’s group’s initial work on this method in 2015 and their proof-of-principle experimental test in 2019.

On the other hand, Hanson says, making quantum repeaters isn’t the only path forward for next-generation quantum network technology.

“In our opinion, simple photon sharing is not the most attractive technology, because many use cases remain out of reach,” says Hanson. “Instead, our goal is to create… entanglement on demand: by combining entanglement delivery via photonic channels with long-lived quantum memories – a buffer of entangled qubits ready for consumption. »

This way, Hanson says, quantum entanglement can be stored like energy in a battery or terabits on a hard drive, and leveraged when users at each end of the network want to share quantum information.

“Buffered entanglement will unlock an exciting range of applications beyond [quantum cryptography] that have the promise of delivering real value,” says Hanson. “It will be interesting to see when Cisco makes the leap to this technology for its networks.”