Researchers in Japan have come up with a way of doing quantum cryptography that could overcome two of the technology's big problems. The new protocol is designed to work with off-the-shelf equipment and use less bandwidth than existing methods. It’s just a mathematical proposal, but it could help make quantum key distribution more commercially viable.
With an encrypted message, the sender and recipient share a key that unscrambles its contents. Ensuring that the key hasn't been stolen is the problem. With quantum cryptography, the key is created at the sender and receiver by transmitting photons over fiber-optic lines. The polarity of a photon—a quantum property that says whether it is oscillating vertically or at angle—can be determined by the receiver and compared with a second "entangled" photon created at the same time. The polarity of the photons is translated into bits that make up a key to decrypt messages.
With quantum key distribution, the security of the transmission is assured by the Heisenberg uncertainty principle. If an eavesdropper tries to intercept the key, it will change the state of the paired photons—an event that can be detected by the sender of the key.
In research published in Nature last week, the Japanese team describes a method for securing communications that doesn’t rely on the uncertainty principle and needs no regular measurement to see if the key's been tampered with.
With this technique, photons are sent over an optical fiber using ordinary lasers, rather than specialized equipment usually needed to create quantum keys. The laser emits a train of photons and a device called a phase modulator imparts a phase on them.
The receiver splits the signal into two separate signals with a randomly generated delay between them. Then those two signals, which are oscillating waves, are superimposed and detected on the receiving end. The combined waves could be out of phase and cancel each out or they could be in phase and create a bigger wave.
The phase difference between pulses can then act as bits that can make up a key to decrypt the message. For example, pulses with the same phase are a bit value of zero, while pulses with a different phase are a bit value of one. When the receiver—who, by convention, is called Bob—detects a photon, he learns whether the superimposed pulses have the same or different phase. Then he tells the sender, called Alice, what the relevant pulse numbers are. Because the sender records all the pulses, she can determine the bit value based on what Bob tells her, explains co-author Masato Koashi from the University of Tokyo.
In an email, Koashi from the University of Tokyo describes how the key is protected from theft by an intruder, called Eve:
One of the keys to securing the communication is to send a large number of optical pulses but they are very weak such that they amount to only a few photons in total. Hence, even if Eve waits for Bob to announce the numbers for two pulses and then measures Alice's signal, the chances of Eve's detecting any photon in the two relevant pulses are very low.
Another key is the fact that Bob generates the delay randomly. Eve may measure Alice's signal immediately and learn the phases of a few pulses. Eve then tries to manipulate Bob's announcement to fall on those pulses for which she has learned the phases. The random delay prevents such a manipulation.
By using this method, the protocol doesn't require regular measurements of the transmissions to discover eavesdroppers as existing quantum key distribution systems do. That significantly reduces the amount of communications overhead required, which is important in situations where the communication channels are noisy, such as when trying to communicate with a satellite during a snowstorm, Koashi says.
The work is still theoretical but an experiment could be done without specialized equipment. “What you need is just to use a conventional laser and a phase modulator, which is already used in digital optical communications,” he says. However, it's still not clear that the performance of the system would be good enough, he adds.
There are already commercial quantum key distribution systems, although their use is fairly limited. Dedicated dark fiber lines are needed and the distance for sending secured data is limited to about 100 kilometers. But there continues to be a significant amount of work in extending the range of quantum cryptography and using shared fiber optic lines.
According to Koashi, researchers in the field expressed surprise with the Nature paper because the security is done by transmitting photons, yet Heisenberg's uncertainty principle isn't relevant. "For them, it's quite a surprise that something different can be done to create secret communications," he says.