Detection efficiency for loophole-free Bell tests with entangled states affected by colored noise

Abstract

Loophole-free Bell tests for quantum nonlocality and long-distance secure communication require photodetection efficiencies beyond a threshold ${\ensuremath{\eta}}_{\mathrm{crit}}$ that depends on the Bell inequality and the noise affecting the entangled state received by the distant parties. Most calculations of ${\ensuremath{\eta}}_{\mathrm{crit}}$ assume that the noise is random and can be modeled as white noise. However, most sources suffer from colored noise. Indeed, since entangled states are usually created as a superposition of two possible deexcitation paths, a partial distinguishability between the two processes leads to the appearance of colored noise in the generated state. Recently, there was a proposal for a loophole-free Bell test [A. Cabello and F. Sciarrino, Phys. Rev. X 2, 021010 (2012)], where a specific colored noise appears as a consequence of the precertification of the photon's presence through single-photon spontaneous parametric down-conversion. Here we obtain ${\ensuremath{\eta}}_{\mathrm{crit}}$, the optimal quantum states, and the local settings for a loophole-free Bell test as a function of the amount of colored noise. We consider three bipartite Bell inequalities with $n$ dichotomic settings: Clauser-Horne-Shimony-Holt ($n=2$), ${I}_{3322}$ ($n=3$), and ${A}_{5}$ ($n=4$), both for the case of symmetric efficiencies, corresponding to photon-photon Bell tests, and for the totally asymmetric case, corresponding to atom-photon Bell tests. Remarkably, in all these cases, ${\ensuremath{\eta}}_{\mathrm{crit}}$ is robust against the colored noise. The present analysis can find application in any test of Bell inequalities in which the dominant noise is of the colored type.