[Paper Review] Quantum and classical telecommunication channel multiplexing

Andreas Lenhard, José Brito+4 more|arXiv (Cornell University)|Jul 22, 2015

Quantum optics and atomic interactions3 citations

TL;DR

This paper demonstrates simultaneous multiplexing of classical laser pulses in the C-band and frequency-converted single photons in the O-band over a single telecom fiber. Using spontaneous parametric down-conversion and quantum frequency conversion, the authors preserve time correlations between heralded quantum and classical signals, achieving low-background, correlated transmission over 20 km with a 43 ns delay shift matching theoretical expectations.

ABSTRACT

We demonstrate the multiplexing of a classical coherent and a quantum state of light in a single telecommunciation fiber. For this purpose we make use of spontaneous parametric down conversion and quantum frequency conversion to generate photon pairs at 854 nm and the telecom O-band. The herald photon triggers a telecom C-band laser pulse. The telecom single photon and the laser pulse are combined and coupled to a standard telecom fiber. Low background time correlation of the classical and quantum signal behind the fiber shows successful telecommunication channel multiplexing.

Motivation & Objective

• To enable coexistence of classical and quantum communication channels in a single fiber using existing telecom infrastructure.
• To overcome high loss in quantum memory-compatible wavelengths (854 nm) by converting to the low-loss O-band via quantum frequency conversion.
• To preserve the temporal correlation between quantum and classical signals through frequency conversion and fiber transmission.
• To demonstrate practical feasibility of quantum network node interconnection using standard fiber and minimal modifications.
• To validate that background noise from cross-talk and detector effects can be mitigated via gated detection and spectral filtering.

Proposed method

• Spontaneous parametric down-conversion (SPDC) generates entangled photon pairs at 854 nm, with one photon frequency-converted to the O-band (1310 nm) using a nonlinear waveguide.
• The 854 nm photon acts as a herald, triggering a 30 ns rectangular laser pulse in the C-band (1550 nm) to serve as a classical clock signal.
• The frequency-converted O-band photon and the C-band laser pulse are combined and coupled into a standard single-mode fiber (SMF-28e).
• Time-correlated detection is performed using superconducting nanowire single-photon detectors (SSPDs) for the O-band and a standard photodiode for the C-band.
• Gated detection is applied to isolate the signal peak from background and cross-talk, with gate timing aligned to the expected coincidence window.
• Fiber dispersion effects are characterized by measuring time delays and pulse broadening between O-band and C-band photons over varying fiber lengths.

Experimental results

Research questions

• RQ1 Can a classical laser pulse and a frequency-converted single photon be successfully multiplexed in a single telecom fiber while preserving their time correlation?
• RQ2 To what extent does fiber dispersion affect the relative timing of O-band and C-band signals, and can this be compensated or accounted for?
• RQ3 Can quantum frequency conversion preserve the temporal correlation of SPDC photons when transmitted through standard fiber?
• RQ4 How does detector dead time and saturation affect the observed coincidence peak shape and visibility?
• RQ5 Can background cross-talk from laser pulses leaking into the single-photon channel be effectively suppressed in practice?

Key findings

• The measured delay shift between O-band and C-band signals was 1.861 ± 0.004 ns/km, matching theoretical predictions based on fiber dispersion.
• The time correlation peak between the heralded O-band photon and the C-band laser pulse was preserved over 20 km of fiber, with a center-of-mass shift of 43 ns, consistent with the expected 37.2 ns delay for that length.
• The coincidence peak width at the base was 34 ns for the 20 km transmission, closely matching the expected value of 32 ns (30 ns laser pulse + 20 km × 108 ps/km).
• A three-peak structure in the coincidence trace was attributed to detector dead time (~10 ns) and saturation effects due to multi-photon content in the laser pulse.
• Cross-talk from laser pulses leaking into the O-band detection channel was identified and isolated via a 244 ns delay fiber, confirming that the main signal peak is distinct and correlated.
• Gated detection successfully suppressed background noise, enabling effective separation of the true signal from accidental coincidences and detector artifacts.