Lateral channels hidden in quantum sources could compromise secure communication

A team of researchers from the University of Toronto Engineering discovered multidimensional side channels hidden in existing quantum communication protocols.

The new lateral channels occur in quantum sources, which are the devices that generate quantum – typical – used to send secure messages. The result could have important implications for quantum safety.

“What makes quantum communication safer than conventional communication is that it uses a property of quantum mechanics called conjugated states”, explains Ph.D. Student Amita Gnanapandithan, principal author on an article published in Physical examination letters.

“For example, the position and the momentum are conjugated variables: when you measure one, you disturb the other. If the two variables are chosen at random for coding, whoever tries to listen to the message will automatically introduce the disturbances that can be detected by the parts that try to communicate.

However, despite its inherent safety, there are still means that quantum communication can be compromised due to imperfections in the devices used for practical implementations.

Between 2000 and 2012, researchers have shown that secondary channels can occur due to how quantum detectors work. These lateral channels act as shortcomings, allowing someone to listen to the signal without introducing a detectable disturbance.

To remedy it, in 2012, Professor Hoi-Kwong Lo and his collaborators developed a new protocol known as the quantum distribution independent of the measure (MDI-QKD). The protocol effectively short-circuit all side channels associated with quantum particle detectors.

With the supported detector, Gnanapandithan, co-supervised by LO and Professor Li Qian, turned to the search for potential secondary channels associated with the other end of communication: source devices.

“Let’s say that you want to code information according to the way in which light from the source is polarized, that we call optical polarization,” explains Gnanapandithan.

“You would use two polarization bases combined for coding, and ideally, you would like to keep your coding in the degree of freedom of polarization. You also do not want polarization to be correlated with any other degree of freedom, because if it is, then the range can measure this second to obtain information on polarization.”

The idea that the degree of freedom of coding is not correlated with other degrees of freedom in quantum optical sources is known as the dimensional hypothesis. Violating this hypothesis means that the message may not be secure.

In practice, today’s quantum sources can often introduce such a violation due, for example, correlations between adjacent signals. This is called the model effect and translates into information on previous signals that flee subsequent signals.

But in the latest study, Gnanapandithan used both theoretical models and quantum physical sources to demonstrate a new source of violation that had never been considered before.

“We knew that the modulation process can be a little distorted, but what we have found is that the modulation process can also be varying over time, even in the same signal optical impulse,” explains Gnanapandithan.

“More specifically, we have made the very subtle awareness that this defect is in fact a violation of the dimensional hypothesis. Consequently, we call this type of multidimensional mudulation hidden from fault,” whose coding varying in time is only an example. “”

The magnitude of these side channels depends on the type of equipment used.

“If your equipment has a higher bandwidth, you can apply a modulation signal to your optical impulse which should bring it closer to what it should ideally be,” she said.

“But if your equipment is seriously limited to the bandwidth, then the modulation impulse could be seriously distorted, which would worsen the problem.

“There is also a new type of quantum key distribution source (QKD) which has been introduced into the literature, called a passive QKD source. Passive QKD sources do not even use modulators, so these bandwidth problems would not apply.”

LO says that the future work of his team will focus on the possible means of reducing the newly discovered secondary channels.

“We can show creativity and perhaps find ways to get around these problems,” he said.

“But as we have learned in the past, it is also possible that our new method can cause its own problems. You never know how many layers there will be, but I think that the very important first step is to simply identify the problems you have to face, and that’s what we have done here.”