Christiana Panayi

Quantum memories are used to improve the ate-versus-distance behavior in measurement-device-independent quantum key distribution (MDI-QKD) systems. The required specifications in terms of reading and writing times for such memories are obtained. It is shown that the faster the access times are, the higher the repetition rates and the lower the required coherence times would be. Additionally this protocol offers an immense security by removing side-attack channels over protocols such as the standard decoy-state BB84 protocol. A comparison of this protocol with the original MDI-QKD is given in terms of secret key generation rate under practical assumptions. Various sources of imperfection such as reading and writing efficiencies of the memories, channel and detector efficiencies and dark count rates are considered. The crossover distance is determined after which the present protocol outperforms the MDI-QKD and a typical quantum repeater network.

Three network architectures, compatible with passive optical networks, for future hybrid quantum-classical networks are proposed and compared. These setups rely on three different schemes for quantum key distribution (QKD): BB84, entanglement-based QKD, and measurement-device-independent QKD (MDI-QKD). It turns out that, while for small-to-moderate-size networks BB84 supports the highest secret key generation rate, it may fail to support large numbers of users. Its cost implications are also expected to be higher than other setups. For large networks, MDI-QKD offers the highest key rate if fast single-photon detectors are employed. Entanglement-based networks offer the longest security distance among the three setups. MDI-QKD is, however, the only architecture resilient to detection loopholes and possibly the most favorable with its less demanding end-user technology.Entanglement-based and MDI-QKD setups can both be combined with quantum repeater systems to allow for long-distance QKD with no trust constraints on the service provider.