Sub-picosecond inter-core skew characterization in multicore fibers via Hong--Ou--Mandel interference

This paper demonstrates a high-precision method for characterizing inter-core skew in multicore fibers using Hong-Ou-Mandel interference, achieving a $\pm0.11$ ps measurement accuracy that significantly surpasses classical techniques and validates the stochastic random-walk scaling of skew over lengths ranging from laboratory to field-deployed scales.

The Gist

Imagine a multicore fiber optic cable as a four-lane highway built inside a single glass tube. In a perfect world, if you send four identical cars (light pulses) down these four lanes at the exact same time, they would all arrive at the destination simultaneously.

However, in reality, the lanes aren't perfectly identical. One lane might have a slightly bumpier road, or a slightly different surface texture. This causes the cars to arrive at slightly different times. This difference in arrival time is called Inter-Core Skew (ICS).

For decades, measuring this tiny time difference in long cables has been like trying to time a race between runners using a stopwatch that only ticks every 10 seconds. It's too slow to catch the split-second differences that matter for high-speed data or quantum computing.

Here is how this paper solves that problem, using some creative analogies:

1. The Problem: The "Bumpy Road" Mystery

The researchers wanted to measure exactly how much slower one lane is compared to another in a commercial 4-lane fiber cable.

• The Old Way: Previous methods were like trying to measure the speed of a car by looking at its blurry headlights from a mile away. They could tell the car was moving, but they couldn't measure the tiny differences in arrival time (only within about 10–20 picoseconds, which is 10 trillionths of a second).
• The Challenge: In long cables, the "bumps" in the road change randomly. If you try to measure them with standard tools, the vibrations of the earth or temperature changes mess up your measurement before you can finish.

2. The Solution: The "Quantum Coincidence" Trick

The team used a clever trick called Hong-Ou-Mandel (HOM) interference. Think of this not as measuring speed, but as listening for a specific "clap" that only happens when two things arrive at the exact same moment.

• The Setup: They sent pairs of "quantum twins" (entangled photons) into the four lanes.
• The Magic: When these twins meet at a special 4-way intersection (a beam splitter), they behave like social butterflies. If they arrive at the intersection at the exact same time, they always leave together through the same exit doors. If they arrive even a tiny bit apart, they split up and go to different doors.
• The Measurement: By adjusting the delay of one lane and watching when the "twins" stop splitting up and start leaving together, the researchers can pinpoint the exact moment the lanes are synchronized.

3. The "Immunity" Superpower

The paper highlights a crucial advantage: This method is immune to noise.

Imagine trying to measure the length of a rope while a strong wind is blowing it around. A standard ruler (classical method) would give you a wrong answer because the rope is moving.
However, the HOM method is like a ghost ruler. It doesn't care if the rope is swaying in the wind; it only cares about the relationship between the two twins. Because of this, they could measure these tiny time differences even in long, installed cables that were vibrating and changing temperature, where other methods would fail.

4. The Results: A New Level of Precision

The team measured cables ranging from a few meters long (like a lab bench) to 1.3 kilometers long (a real-world field cable).

• The Precision: They achieved a precision of ±0.11 picoseconds. To put that in perspective, if the old methods were like measuring a race with a stopwatch that has a 10-second error, this new method is like measuring with a stopwatch that has an error smaller than the blink of an eye. It is about 180 times more precise than the current standard.
• The Discovery: They confirmed that as the cable gets longer, the "bumpiness" (skew) doesn't just add up in a straight line. Instead, it grows like a random walk. Imagine a drunk person walking down a hallway; they don't walk in a straight line, but their distance from the start grows with the square root of the steps taken. The researchers proved this "random walk" model holds true from the lab bench all the way to a 1.3-kilometer field cable.

5. Why It Matters (According to the Paper)

The paper states this technology is a practical platform for two main things:

1. Classical Internet: It helps ensure that data sent through different lanes of a fiber cable arrives in sync, which is vital for next-generation super-fast internet.
2. Quantum Networks: It allows scientists to fix timing mismatches before they ruin delicate quantum experiments, ensuring that "quantum twins" can still talk to each other even after traveling through long, imperfect cables.

In short: The researchers built a super-precise "quantum stopwatch" that can measure the tiny timing differences between lanes in a fiber optic cable, proving that these differences grow in a predictable, random pattern, and doing it with a level of accuracy that was previously impossible for long cables.

Technical Summary

Technical Summary: Sub-picosecond Inter-core Skew Characterization in Multicore Fibers via Hong–Ou–Mandel Interference

Problem Statement
Inter-core skew (ICS), defined as the differential group delay between signals co-propagating in different cores of a multicore fiber (MCF), is a critical limiting parameter for both classical space-division multiplexed (SDM) communications and quantum photonic networks. In classical systems, ICS necessitates larger digital signal processing buffers and degrades carrier-phase estimation. In quantum protocols involving broadband photon interference (e.g., QKD, entanglement swapping), ICS acts as a source of decoherence.

Existing characterization methods, such as Correlation Optical Time-Domain Reflectometry (C-OTDR) and white-light interferometry, typically achieve resolutions of only 10–20 ps. While sufficient for long-haul classical links, this precision is inadequate for sub-picosecond requirements in quantum applications and short integrated photonic segments. Furthermore, ICS does not accumulate deterministically; it is dominated by stochastic, spatially-varying fluctuations. Previous attempts to engineer away ICS via splice rotations have failed, confirming the need for precise measurement rather than mitigation. Classical interferometric methods are also impractical for long, installed fibers due to their sensitivity to first-order path fluctuations, which require active phase stabilization.

Methodology
The authors present a high-precision measurement technique utilizing two-photon Hong–Ou–Mandel (HOM) interference within a fiber-integrated $4 \times 4$ multiport beam splitter. The experimental approach involves two distinct regimes to cover different fiber lengths:

1. Laboratory Regime (7.7 m – 187.7 m): Photon pairs are generated via spontaneous parametric down-conversion (SPDC) in a PPKTP crystal pumped by a 775 nm mode-locked laser. These degenerate 1550 nm pairs are coupled into a commercial four-core fiber (Fibercore SM-4C1500). The photons traverse the fiber and interfere in a $4 \times 4$ MCF-based beam splitter implementing a Hadamard transformation. Coincidence counts are detected by superconducting nanowire single-photon detectors (SNSPDs).
2. Field-Deployed Regime (1300 m): For longer fibers where channel losses (dominated by connectors and splices) render SPDC coincidence rates impractical, a bimodal pulsed laser source (300 ps pulses at 1550 nm) is used. The measurement relies on the two-photon component of the coherent pulse. The HOM interference pattern is modulated by a $\cos^2$ beat structure due to the laser's multi-mode longitudinal structure, allowing for precise extraction of the delay center despite the broader pulse width.

Key Contributions and Results

• Sub-picosecond Precision: By extracting the center positions of HOM interference dips and peaks across all twelve input core-pair combinations, the authors demonstrate a measurement precision of $\pm 0.11$ ps. This represents an approximately 180-fold improvement over the standard C-OTDR method. The precision is currently limited by the delay-stage positioning uncertainty ($\sim 0.15$ ps) rather than photon statistics.
• Stochastic Scaling Validation: The study provides the first direct validation of the stochastic random-walk scaling of ICS across a length range spanning laboratory to field-deployed scales (7.7 m to 1300 m). The root-mean-square (RMS) ICS, $\sigma_\tau(L)$, follows the relationship:
  $$\sigma_\tau(L) = \kappa\sqrt{L} + c$$
  where $\kappa = 48.7 \pm 2.5$ ps/$\sqrt{\text{km}}$ and $c = 9.76 \pm 1.2$ ps. The coefficient $\kappa$ is significantly larger than typical polarization mode dispersion (PMD) coefficients in single-mode fibers, confirming the effect is genuine inter-core differential group delay.
• Robustness to Fluctuations: The HOM method is highlighted for its immunity to first-order path fluctuations. Unlike classical interferometry, the coincidence dip depends only on the second-order temporal envelope, making it robust for measurements in long, installed fibers without active phase stabilization.
• Fundamental Limits: A Fisher information analysis establishes a fundamental Cramér–Rao precision limit in the femtosecond range (0.93–1.94 fs per channel), indicating that the current $\pm 0.11$ ps precision is two orders of magnitude above the theoretical quantum limit and could be improved with better delay control.

Significance
The paper claims that these results establish a practical platform for characterizing timing uniformity in MCF-based networks. For classical SDM, the demonstrated precision exceeds digital signal processing buffer requirements, enabling per-span ICS monitoring compatible with deployed telecom infrastructure. For quantum networks, the ability to characterize ICS with sub-picosecond resolution allows for the compensation of timing mismatches before they degrade photon interference. The successful extension of this characterization to a 1300 m field-deployed fiber using a pulsed laser source demonstrates the method's viability for real-world, high-loss environments where classical interferometric methods fail. The work confirms that ICS accumulation is stochastic and validates the $\sqrt{L}$ scaling law over a broad length spectrum, a feat previously unachieved with direct multi-length fitting.