Realistic photon-number resolution in generalized Hong-Ou-Mandel experiment

TL;DR

This paper derives analytical expressions for photocounting distributions in a generalized Hong-Ou-Mandel experiment using realistic photon-number resolving detectors, highlighting how detector imperfections affect nonclassical photon statistics.

Contribution

It introduces a model for realistic PNR detectors in a multimode Hong-Ou-Mandel setup and shows that postselected event probabilities relate to ideal detector cases.

Findings

Probabilities are proportional to those with perfect detectors

Analytical formulas for realistic detector effects

Illustrations with on/off detectors and dead time effects

Abstract

We consider realistic photodetection in a generalization of the Hong-Ou-Mandel experiment to the multimode case. The basic layout of this experiment underlies boson sampling -- a promising model of nonuniversal quantum computations. Peculiarities of photocounting probabilities in such an experiment witness important nonclassical properties of electromagnetic field related to indistinguishability of boson particles. In practice, these probabilities are changed from their theoretical values due to the imperfect ability of realistic detectors to distinguish numbers of bunched photons. We derive analytical expressions for photocounting distributions in the generalized Hong-Ou-Mandel experiment for the case of realistic photon-number resolving (PNR) detectors. It is shown that probabilities of properly postselected events are proportional to probabilities obtained for perfect PNR detectors. Our results are illustrated with examples of arrays of on/off detectors and detectors affected by a finite dead time.