Quantum State Characterization Using Measurement Configurations Inspired by Homodyne Detection

TL;DR

This paper explores how different measurement configurations inspired by homodyne detection can be used to characterize unknown quantum optical states, including theoretical analysis and an experimental demonstration involving polarization-dependent beam splitter actions.

Contribution

It introduces a novel measurement scheme for quantum state characterization based on photon counting with configurable local oscillator parameters and different beam splitter actions.

Findings

The measurement configuration allows inference of the unknown state parameters.

Varying LO phase and intensity reduces the number of detectors needed.

Experimental validation demonstrated the theory with polarization-dependent beam splitters.

Abstract

In the standard homodyne configuration, an unknown optical state is combined with a local oscillator (LO) on a beam splitter (BS). Good quadrature measurements require a high-amplitude LO and two high-efficiency photodiodes whose signals are subtracted and normalized. By changing the LO phase, it is then possible to infer the optical state in the mode matching the LO. For quantum information processing, the states of interest are in well-separated modes, corresponding to a pulsed configuration with one relevant LO mode per measurement. We theoretically investigate what can be learned about the unknown optical state by counting photons in one or both outgoing paths after the BS, keeping the LO mode fixed but choosing its phase and magnitude. We consider measurement configurations where the BS acts differently on different sets of matching modes. When the BS acts identically on all matching modes it is possible to determine the content of the unknown optical state in the mode matching the LO conditional on each number of photons in the orthogonal modes on the same path. In particular, if both the phase and the intensity of the LO can be varied, then the statistics of just one of the counters is enough to infer these parameters, while in the case of an LO with fixed intensity both detectors are needed to accomplish this. Our results are derived by demonstrating a bijection, or lack thereof, between the probability distributions over the space of outcomes of the counter(s) and certain parameters of the unknown state for different measurement configuration. We report an experiment that was conducted to demonstrate the theory in the case where the BS acts differently depending on the polarization.