In the internet one faces completely different challenges than tossing coins...
In today’s communication networks, many use cases exist in which the communicating parties interact remotely on the distance, but do not trust each other. Examples are electronic voting procedures, contract signing or in general authentication procedures. The coin toss problem is important for all these cryptographic tasks: In fact, it is a fundamental building block for many complex cryptographic protocols, something that can be understood as (construction) manuals for data encryption. It’s like Lego: If you follow a sequence of steps in the manual, you’ll get a reproducible result. In data communication, mathematical or physical operations lead to securely encrypted messages.
There are already many cryptographic protocols, aren’t there?
Exactly, including those that use quantum physics to achieve absolute, physical security in data communication. A prominent example is quantum key distribution, which my team also conducts a lot of research on. This method enables two parties to generate a secret key that is known only to them. This is perfect to secure the data exchange between trusted parties.
And this is not enough?
No, because once the parties do not trust each other, this protocol works no longer. This motivated us to realise the quantum coin toss experimentally with single photons.
Earlier we talked about an ordinary coin toss based on classical physics. But what’s a quantum coin toss?
Let’s start first with explaining a classical coin toss in the cryptographic context, which also works without quantum technology. Imagine a scenario in which two people at the distance want to agree remotely on one or several random bit values - zero or one, heads or tails - without involving a third party. Compared to an ordinary, physical coin toss, there are obviously two decisive boundary conditions: The two persons are separated by a large distance and they cannot involve a third person as a ’referee’. Such a coin toss realised over the distance using purely computational, classical protocols will not work, as soon as one or both parties have sufficient computing power to cheat.
And a quantum coin toss works?
At least it is much safer. In the quantum version, where qubits are used instead of classical bits, a coin toss can be realised over the distance more securely than in the classical world. In our experiment, we use single photons instead of coins, with the polarisation state of the photons corresponding to heads or tails. The sender ’flicks’ a single photon with random polarisation towards the receiver, using for example an optical fiber. The receiver measures the polarisation state and hence "sees" whether heads or tails is facing up.
The light source is crucial for the experiment, right?
Indeed! For our realization of quantum coin tossing, or flipping, we used for the first time a quantum light source emitting single photons at the push of a button. Previous quantum methods use attenuated lasers or other types of so-called statistical quantum light sources. Using our method, we were able to make the quantum coin toss even more secure. The effect is still small, but we already have ideas on how to improve it further.
Original publication
Daniel A. Vajner, Koray Kaymazlar, Fenja Drauschke, Lucas Rickert, Martin von Helversen, Hanqing Liu, Shulun Li, Haiqiao Ni, Zhichuan Niu, Anna Pappa, Tobias Heindel (2026): Single-Photon Advantage in Quantum Cryptography Beyond QKD. Nature Communications 17, 2074; DOI: 10.1038/s41467-026-69995-9