Quantum key distribution (QKD) offers information-theoretical security for two communicating parties, Alice and Bob. However, while theoretically secure, a QKD system can suffer from security loopholes (“side-channels”), particularly in detectors. For instance, hackers can send in a strong light to force wrong detector clicks (called a “detector blinding attack”) or exploit the small unbalance in detection time windows of pairs of detectors (called a “time-shift attack”).
To address these vulnerabilities, researchers have proposed the measurement-device-independent (MDI) QKD protocol and its variants. In these protocols, instead of Alice sending a quantum signal to Bob, they both send signals to an untrusted third party Charlie, who measures the signals and publicly announces the results. In this protocol, Alice and Bob can successfully generate a pair of secure keys without Charlie being able to gain any knowledge of the keys. Therefore, Charlie is allowed to be an untrusted party, and the protocol is automatically immune to any side-channels in his detectors.
However, a major challenge that MDI QKD protocols have is that, since a third party Charlie is involved (and a physical phenomenon called photon interference is used at Charlie), the photonic loss in the pair of channels Alice-Charlie and Bob-Charlie are required to be *symmetric* in order for the protocol to work properly with good performance. In reality, though, due to geographical locations or moving platforms, the pairs of channels are seldom naturally symmetric, which greatly limits the application of MDI-QKD protocols over the years.
We address this problem from the software side and designed a new type of *asymmetric* MDI-QKD protocols [1-4] which automatically compensates for asymmetric channel pairs by adjusting the source signal intensities, enabling a good performance of MDI-QKD protocols even when channels are asymmetric. This allows us to build scalable and multi-user quantum communication networks based on MDI-QKD, greatly improving the practical security of QKD networks. In the future, we are interested in designing and optimizing more types of MDI protocols and their variants that are suitable for large-scale quantum networks.
[1] W Wang, F Xu, HK Lo. "Asymmetric protocols for scalable high-rate measurement-device-independent quantum key distribution networks." Physical Review X 9.4 (2019): 041012.
[2] H Liu, et. al. "Experimental demonstration of high-rate measurement-device-independent quantum key distribution over asymmetric channels." Physical Review Letters 122.16: 160501 (2019).
[3] W Wang, HK Lo. "Simple method for asymmetric twin-field quantum key distribution." New Journal of Physics 22.1: 013020 (2020).
[4] X Zhong, W Wang, HK Lo, L Qian. "Proof-of-principle experimental demonstration of twin-field quantum key distribution over optical channels with asymmetric losses." npj Quantum Information 7, 8 (2021).