Clock Synchronization with Weakly Correlated Photons

Clock synchronization is necessary for communication and distributed computing tasks. Previous schemes based on photon timing correlations use pulsed light or photon pairs for their strong timing correlations. In this work, we demonstrate successful synchronization of crystal clocks using weakly time-correlated photons of 180 ns coherence time from a bunched light source. A synchronization timing jitter of 10 ns is achieved over symmetric -102 dB optical channel loss between two parties, over a span of 25 hours. We also present a model that gives better estimates to the coherence peak finding success probabilities under low signal.

💡 Research Summary

The paper presents a novel approach to clock synchronization that relies on weakly time‑correlated photons rather than the strong correlations typically obtained from pulsed lasers or spontaneous parametric down‑conversion (SPDC) photon pairs. The authors generate a “bunched” light source by sending a 780 nm laser through an unbalanced Mach‑Zehnder interferometer, producing a second‑order coherence function g^(2)(τ)=1+½ exp(−2|τ|/τ_c) with a measured coherence time τ_c = 180 ns and a modest peak value g^(2)(0)=1.42. This source emits photons with only weak temporal correlations (g^(2)≤2), yet the correlation peak is still detectable after transmission through two independent optical channels each incurring roughly –102 dB of loss.

In the experimental setup, the bunched light is split and sent to two remote stations. Each station uses a silicon avalanche photodiode (Si‑APD) and a time‑tagger disciplined by an independent free‑running 10 MHz quartz crystal oscillator. The single‑photon detection rates are about 190 kcps per side, and the coincidence window is set to 256 ns. Over a 25‑hour run the system records an average of 8.5 kHz true coincidences against an accidental background of ~7 kHz, yielding a clear cross‑correlation peak that can be tracked in real time.

The core technical contribution lies in the statistical model for peak‑finding success. Traditional analyses approximate the noise in the fast Fourier transform (FFT) cross‑correlation by a Gaussian distribution, which becomes inaccurate when the signal‑to‑noise ratio is low. Here the authors treat the number of accidental coincidences in each time bin as a Poisson variable with mean λ = s₁ s₂ δt T, where s₁ and s₂ are the single‑photon rates, δt the bin width, and T the acquisition time. The maximum count across N bins follows a max‑order Poisson distribution, allowing an exact expression for the probability that the true coincidence peak exceeds the largest accidental bin. They derive a condition for successful peak identification: c_e > ξ ν T ·