From ATM withdrawals to military secrets, sensitive information that travels over the internet is encoded so hackers cannot see it. The encryption algorithms that protect it are so powerful, theorists claim it would take 50 supercomputers longer than the age of the universe to guess how to crack them. Yet even encrypted internet traffic remains vulnerable. Some algorithms are simply not as strong as previously thought.
More often, though, it is the very nature of the public internet that is the weak point. For two parties to encrypt and decrypt information, they must share a digital key to lock and unlock encrypted data. Anyone who listens in on internet traffic and steals the key and the data can turn secrets into open books. Fortunately, there is a solution that uses the principles of quantum physics to create an internet connection that no one can spy upon.
The outline of one such secure network is taking shape at the Kavli Institute of Nanoscience at Delft Technical University in The Netherlands.
“The security of quantum networks is based on quantum entanglement, a special property that two quantum bits can have,” explained Stephanie Wehner, a physicist and roadmap leader of the Quantum Internet and Networked Computing initiative at QuTech, a leading quantum computing initiative. Like many core QuTech researchers, Wehner is also a member of Delft’s Kavli Institute.
Quantum entanglement describes a tight connection between two particles, such as electrons or photons of light. These particles, called qubits, are basic units of quantum information.
The properties of entangled qubits make them ideal for sharing information secretly. First, they are monogamous: no more than two qubits can share an entanglement. This makes their connection fundamentally unhackable by an outsider, Wehner said. Second, entangled qubits behave exactly like one another. Take a measurement on one and a measurement on the other always yields the exact same result. Albert Einstein once referred to this as “spooky action at a distance,” since the particles appeared linked even when separated. This makes it possible for two parties to measure a series of entangled photons and turn those measurements into a string of numbers only the two of them could possibly know. Both parties could then use those numbers to create an encryption key without ever storing or transmitting the key over a public internet connection.
The concept is simple, but executing it is hard. For one thing, Wehner said, quantum communications is noisy. Sometimes, measurements are imprecise. Other times, the fragile quantum state of entangled photons collapses as they travel through optical fibers.
One way to send a stream of qubits, measure some of them, and then run a statistical test on the measurements. If the data looks good, it would let the two parties create an algorithm to correct deviations in measurements. If there are too many errors, it might indicate that someone is trying to intercept those qubits.
Distance presents another problem. Transmitting entangled qubits more than 50 km is difficult. “We’re sending one single particle of light down an optical fiber,” Wehner said. “It is very easily absorbed and lost if the fiber is too long.”
To go longer distances, QuTech is developing a quantum repeater. The internet now uses repeaters to boost signals that weaken during transmission. Entangled qubits, however, cannot be amplified like a classical stream of electrons. Instead, Wehner explained, quantum repeaters have to “swap” the entanglement of one qubit with another. “The way you can think of swapping is that it glues two entangled segments together to form one larger, long-distance entanglement,” Wehner said.
Entanglement swapping, however, takes time. This is because a repeater must be able to store two qubits in memory long enough to execute the operation. This is still a work in progress. Today, repeaters are very slow, able to process only a few qubits per second. For a true quantum network, they would need to run thousands of times faster.
Meanwhile, QuTech plans to build a test network between nodes in Delft and The Hague within the next two years, Wehner said. A simplified repeater, one without quantum memory, will entangle qubits and send them to the nodes to prove it can create a connection. Ultimately, Wehner expects that to grow into a larger and more robust network where entangled qubits deliver information as well as security codes and link quantum computers to solve some of science’s toughest problems.