Coincidence detection for photon triplet sources

Photon triplet generation based on third-order spontaneous parametric down-conversion remains as an experimental challenge. The challenge stems from the trade-offs between source brightness and instrument noise. This work presents a probability theory of coincidence detection to address the detection limit in source characterization. We use Bayes' theorem to model instruments as a noisy communication channel and apply statistical inference to identify the minimum detectable coincidence rate. A triplet generation rate of 1-100 Hz is required for source characterization performed over 1-72 hours using superconducting nanowire single-photon detectors.