5 questions to Prof. Philippe Grangier, academic leader of the QKISS European project led by Exail

How did your research career bring you from the manipulation of individual photons to quantum-secure communication?

I began research in 1980, working with Alain Aspect on the famous experimental tests of Bell’s inequalities. Until 1986, I pursued my PhD thesis on the preparation and manipulation of single-photon states of the light, which included the first demonstration of single-photon interferences. As you can imagine, this was a great introduction to quantum optics.

Grangier’s single photon interferences experiment, exhibited at the small museum of Institut d'Optique Graduate School.

After completing my PhD, I carried out a postdoc in Bell Labs in New Jersey, with Dick Slusher, who had just observed squeezed states of light for the first time in 1985. In the stimulating environment of Bell Labs, I set up a new experiment and demonstrated in 1987 that squeezed states - quantum states of light with controlled fluctuations - can improve the sensitivity of interferometric measurements. This led me to study the ultimate limits of quantum measurements, in particular what are called Quantum Non-Demolition (QND) measurements. Returning to France in 1988, I set up a series of experiments to realize different kinds of QND measurements, using methods from non-linear optics (P. Grangier et al, Nature, 1998).

At the same time, interest in quantum technologies was growing, and European research programs were launched on quantum computing and quantum-secure communication. I became directly involved in several projects, including one on quantum-secure communication (QKD). Interestingly, our earlier work on QND measurements could be used to calculate the limits on what an eavesdropper can learn in a QKD scheme if they attempt to intercept a secret key encoded on a weak laser beam. This work with my PhD student Frédéric Grosshans gave rise to a new QKD protocol, now called GG02 (Grosshans-Grangier 2002). It is not (yet) as famous as the very first QKD protocol, BB84 (Bennett-Brassard 1984), but it is gaining traction. We demonstrated our protocol experimentally in 2003 (F. Grosshans et al, Nature, 2003).

F. Grosshans et al. Quantum key distribution using gaussian-modulated coherent states, Nature (2003)

Then, there were many further developments - better security proofs, improved experiments – carried out in collaboration with Thierry Debuisschert (Thales) and Eleni Diamanti (CNRS / LIP6). This work was supported by a series of European projects, including SECOQC, with culminated in a full demo in Vienna in 2008.

What is your perspective as a physicist on the different QKD schemes? Are there some determining factors when choosing between CV-QKD and DV-QKD systems?

Our GG02 protocol uses continuous variables QKD (CV-QKD), whereas BB84 relies on discrete variables QKD (DV-QKD). All QKD protocols exploit the fact that some quantum variables do not commute - measuring one disturbs the other. The secret key is randomly encoded in one variable or the other, but the choice is unknown in advance by the eavesdropper, Eve. Any attempt by Eve to intercept the transmission introduces errors or noise when she tries to measure the wrong variable. After further processing, this allows the legitimate users, Alice and Bob, to extract a secret key that is error-free and unknown to Eve – provided the initial error rate is sufficiently low. Otherwise, no key is established. In DV-QKD, the key is encoded by Alice in non-commuting photon properties, such as different directions of polarization, and Bob is using photon counting detectors.

In CV-QKD, by contrast, the key is encoded on complementary variables that cannot be measured simultaneously, such as amplitude and phase of a small laser pulse. DV-QKD requires photon counters, which are delicate and not widely used in optical telecommunications. CV-QKD, on the other hand, uses only standard telecom technologies. In fact, a CV-QKD system is essentially a high-performance, low noise coherent optical telecommunications system. This is why CV-QKD has become increasingly popular, with several startups entering the field.

What are the main technical challenges facing the industrial deployment of QKD systems?

A critical issue is that QKD security is guaranteed only if the device is properly implemented and free from « side channels » that could compromise security in practice. This is true for classical cryptographic devices as well: identifying and eliminating side channels that leak secrets is essential. Over the last 20 years, both QKD devices and security proofs have advanced significantly, but there is now a strong need for standardization and certification, similar to what exists for classical cryptography. Active work is ongoing in this area for both CV and DV approaches, for instance within the European project NOSTRADAMUS, a testing infrastructure for quantum key distribution.

It is also important to note that CV-QKD systems rely on three main technological building blocks: lasers, modulators, and electronics for detection and data processing. Since QKD concerns secure communications, sovereignty issues are critical for each of these components. European countries investing in QKD want to ensure trusted supply chains for these technologies.

How did you start collaborating with Exail on QKD?

We began collaborating with Exail around 2020, within the framework of the European project OpenQKD, the precursor to QKISS. Exail has recognized industrial expertise in manufacturing electro-optical modulators as well as in optical integration. This unique combination of cutting-edge optical technology, integration, and industrialization capabilities is rare in France and perfectly suited our goals.

Moreover, Exail has a long track record of collaborating with research institutions, including participation in multiple European projects. Finally, as mentioned earlier, having a 100% French supplier is crucial in QKD, given the sovereignty concerns.

QKISS demonstrator is composed of Alix emitter channel and Beatrix receiver channel, connected by an optical fiber link.

What role could the QKISS system play in Europe’s QKD landscape?

The ultimate goal of QKISS is to provide a CV-QKD system that delivers a fully European solution for France, deployable on a European QKD network alongside other systems. Significant progress has been made recently in ensuring interoperability between different QKD systems, both DV and CV, through the use of Key Management Systems (KMS).

Looking towards this Quantum Future, I wish success and long life to Alix and Beatrix!