Researchers at ETH Zürich and National University of Singapore have carried out a study investigating whether advantage distillation, a classical cryptography technique that has so far never been successfully implemented, can be applied to device-independent quantum key distribution (DIQKD) systems with the aim of creating a secret key for communication between different parties.
DIQKD protocols are an adaptation of more traditional quantum key distribution (QKD) approaches. They are based on past observations suggesting that when these correlations violate a Bell inequality, a secure key can be extracted even if the different users devices are not fully characterized.
In other words, when assessing the security of DIQKD protocols, users do not need to assume that communicating devices are operating according to their specifications. This is in stark contrast with the device dependency observed in traditional QKD protocols, which typically assume that connected devices are implementing a specific range of quantum operations.
This unique characteristic of DIQKD can significantly enhance the security of communications and data exchanges, as it remains secure even if an attacker manages to influence the behavior of the users devices. This increased security, however, is often accompanied by a crucial limitation: To achieve positive keys rates, DIQKD protocols require low noise levels.
The team tried to overcome this limitation using a cryptography technique known as advantage distillation. This technique allows two people who are communicating to identify segments of cosmic background radiation where they have an advantage over an intruding party.
In DIQKD, however, the radiation is replaced by a signal consisting of entangled particle pairs, distributed by an untrusted source, which may even be controlled by the third, intruding party. Based on this similarity, the researchers set out to explore whether the idea of advantage distillation is actually applicable to DIQKD and whether it can improve its tolerance against noise.
The paper has been published in Physical Review Letters.