Mesoscopic quantum cryptography

Abstract

Since a strictly single-photon source is not yet available, in quantum cryptography systems, one uses, as information quantum states, coherent radiation of a laser with an average number of photons of μ ≈ 0.1–0.5 in a pulse, attenuated to the quasi-single-photon level. The linear independence of a set of coherent quasi-single-photon information states leads to the possibility of unambiguous measurements that, in the presence of losses in the line, restrict the transmission range of secret keys. Starting from a certain value of critical loss (the length of the line), the eavesdropper knows the entire key, does not make errors, and is not detected—the distribution of secret keys becomes impossible. This problem is solved by introducing an additional reference state with an average number of photons of μ{sub cl} ≈ 10{sup 3}–10{sup 6}, depending on the length of the communication line. It is shown that the use of a reference state does not allow the eavesdropper to carry out measurements with conclusive outcome while remaining undetected. A reference state guarantees detecting an eavesdropper in a channel with high losses. In this case, information states may contain a mesoscopic average number of photons in the range of μ{sub q} ≈ 0.5–10{sup 2}. The protocol proposed is easy to implement technically, admits flexible adjustment of parameters to the length of the communication line, and is simple and transparent for proving the secrecy of keys.