Abstract:

While quantum key distribution (QKD) is possible without producing entangled states, there is an advantage in actually preparing an entangled pair of pulses and distributing them to two receivers.  The advantage is the fact that we can totally forget about the credibility of the light source: any fault in or any malicious tampering with the light source will be automatically caught as an increase of observed error rates.  This is certainly true if we use ideal photon detectors, and a nontrivial question is whether we can still enjoy this auto-detecting capability with practical detectors, which cannot distinguish photon numbers or selecting out a single optical mode.  In this talk, I will discuss two independent  approaches: (1) Using a security proof based on complementarity, it can be shown that if one of the receivers uses an ideal detector, the other one can use almost any kind of detectors.  (2) By analyzing properties of polarizations of photons, we can show that auto-detecting capability is available when the receivers use practical detectors modeled by a threshold detector.