Demonstration of MIMO-DFS over 100km of unamplified SSMF Link using Active Laser Drift Stabilization and Optimized Probing Codes

This paper demonstrates reduced noise floor in a 100km unamplified SSMF link for coherent distributed fiber sensing by employing active laser drift stabilization and optimized probing codes to mitigate laser frequency noise.

The Gist

The Big Picture: Turning Phone Cables into Giant Ears

Imagine the thousands of miles of fiber optic cables already buried underground, carrying your Netflix streams and Zoom calls. These cables are usually just "pipes" for data. This paper is about turning those pipes into giant, super-sensitive ears that can listen to the ground, detect earthquakes, or hear a truck driving by, all without digging up the road or adding new cables.

The researchers wanted to make these "ears" work over very long distances (100+ kilometers) without needing expensive signal boosters. They succeeded by fixing a specific problem: the laser source was too jittery.

---

The Problem: The "Shaky Hand" of the Laser

To "listen" through a fiber optic cable, you shine a laser beam into it and listen for the tiny bit of light that bounces back (like shouting in a canyon and hearing the echo).

However, to get high-quality sound over 100km, you can't just shout a single word. You have to send a complex, long, coded message.

• The Analogy: Imagine trying to hear a whisper from someone 100km away. If you shout a simple "Hello," the echo gets lost in the wind. But if you shout a complex, rhythmic song, the echo is easier to pick out.
• The Issue: To hear this "song" clearly, the laser needs to be perfectly steady. But real lasers are like drunk singers. As time goes on, their pitch (frequency) wobbles slightly.
• The Result: Because the laser's pitch is wobbling, the "echo" gets scrambled. It's like trying to recognize a song when the singer keeps changing the key randomly. The background noise gets so loud that you can't hear the tiny vibrations you are looking for.

The researchers discovered that this "wobble" is worst at low frequencies (slow, lazy wobbles). It's not the fast, tiny shakes that ruin the signal; it's the slow, drifting pitch changes that happen over a few seconds.

The Solution: The "Tuning Fork" Stabilizer

To fix this, they built a special device called an Optical Frequency Discriminator (OFD).

• The Analogy: Think of the laser as a musician playing a note. The OFD is a perfectly tuned tuning fork sitting right next to them.
  • If the musician's note drifts even a tiny bit, the tuning fork "screams" (sends a voltage signal) saying, "Hey! You're flat!"
  • A computer (the servo controller) hears this and instantly tells the musician to adjust their pitch back to the right note.
• The Magic: This happens thousands of times a second. It locks the laser's pitch so tightly that the "drunk singer" becomes a "perfectly steady choir."

The Experiment: The 100km Test

The team tested this in a lab using a spool of fiber optic cable that was 123 kilometers long (about 76 miles).

1. Before the Fix: They tried to listen for a vibration (a piezo actuator shaking the cable) from 100km away. The "noise floor" (the static hiss) was so loud that the vibration was completely hidden. It was like trying to hear a pin drop in a rock concert.
2. After the Fix: They turned on the "Tuning Fork" stabilizer.
   • The Result: The static hiss dropped dramatically (by more than half).
   • The Victory: Suddenly, the 120 Hz vibration they created was clearly visible, standing out 10 times louder than the background noise. They could detect a tiny 180-nano-strain change (imagine stretching a rubber band by the width of a hair) over a massive distance.

Why This Matters

This is a game-changer for two reasons:

1. No New Cables Needed: We can use the existing "dark" fibers in telecom networks to monitor earthquakes, landslides, or construction work.
2. No Expensive Boosters: Usually, listening over 100km requires adding expensive amplifiers (like repeaters for a phone line). This method works unamplified, meaning it's cheaper and easier to install.

Summary

Think of the fiber optic cable as a long, thin guitar string.

• The Old Way: You pluck the string, but the air is windy and the string is vibrating on its own, so you can't hear the specific note you plucked.
• The New Way: The researchers built a "wind shield" (the laser stabilizer) that stops the air from blowing and a "tuner" that keeps the string perfectly tight.
• The Outcome: Now, they can hear the faintest pluck from 100 kilometers away, turning the entire global internet infrastructure into a massive, sensitive earthquake detector.

Technical Summary

1. Problem Statement

The paper addresses the challenge of extending the reach and sensitivity of Distributed Fiber Sensing (DFS) systems over existing optical telecom infrastructure without using optical amplifiers (like Raman amplification) or specialized chirped pulse techniques.

• Core Limitation: While Multi-Input Multi-Output (MIMO) DFS using polarization diversity and digital coding offers benefits like mitigated polarization fading and compatibility with live WDM networks, its performance over long distances (e.g., >100 km) is severely degraded by laser frequency noise.
• Specific Mechanism: In time-spreading interrogation methods (where the modulation period $T_p$ is comparable to or longer than the fiber round-trip time $T_r$), slow fluctuations in the laser's central frequency corrupt the channel estimation of the backscattered signal.
• Gap: Previous models often assumed an ideal Lorentzian laser noise profile. However, real-world high-quality sensing lasers exhibit increased frequency noise at low frequencies ("1/f" noise region), which significantly impacts DFS sensitivity in a spectral zone that was not previously precisely characterized for long-reach systems.

2. Methodology

The authors employed a combination of advanced simulation, theoretical modeling, and experimental validation to solve the problem.

A. Simulation and Modeling

• System Model: They simulated a MIMO-DFS setup using a PDM-BPSK probing code over a 50 km emulated fiber.
• Noise Analysis: They introduced a method to quantify the impact of laser frequency noise by integrating the noise floor over the fiber distance. They applied digital high-pass and low-pass filters to the laser frequency noise Power Spectral Density (PSD) to isolate the specific frequency bands contributing to the noise floor.
• Laser Characterization: Instead of relying solely on a standard Lorentzian model, they utilized the measured frequency noise PSD of a real Teraxion NLL laser. This revealed that the "worst-case" noise impact occurs at lower frequencies (around 550 Hz) compared to the idealized model (2 kHz).

B. Active Laser Stabilization

• Hardware Solution: To mitigate low-frequency noise, they designed a compact Optical Frequency Discriminator (OFD) system (based on Silentsys technology).
• Control Loop: The OFD converts laser frequency fluctuations into voltage fluctuations. These are fed into a servo controller that drives the frequency modulation port of the Teraxion laser.
• Target Band: The system was optimized to stabilize the laser specifically in the critical low-frequency range (10 Hz – 10 kHz) where DFS sensitivity is most vulnerable.

C. Experimental Setup

• Link: A 123 km Standard Single-Mode Fiber (SSMF) spool was used to simulate a full telecom span without amplification.
• Probing Code: An optimized PDM-MPSK code was used (217 symbols, 100 Mbaud, 1 dBm output power).
  • This created a code duration ($T_p$) of 1.31 ms and a native gauge length of 1 meter.
  • The repetition rate allowed for a mechanical sensing bandwidth of 382 Hz.
• Validation: They measured the phase standard deviation (noise floor) along the fiber and introduced a controlled mechanical event (a 120 Hz sine wave excitation via a piezo actuator) at the 101 km mark to test detection capability.

3. Key Contributions

1. Precise Spectral Characterization: The paper is the first to demonstrate that DFS noise floor is primarily induced by slow variations of the light source. It precisely maps the "spectral zone of impact" (specifically 100 Hz – 2.5 kHz for their setup) based on the fiber round-trip time, code duration, and the specific laser's noise PSD.
2. Active Stabilization System: Development and deployment of a compact, high-bandwidth OFD-based stabilization loop that reduces laser frequency noise by a factor of $10^3$ in the critical low-frequency band (<10 kHz).
3. Long-Reach Unamplified Demonstration: Successful demonstration of MIMO-DFS over 100+ km of unamplified fiber, achieving 1-meter spatial resolution without Raman amplification.

4. Key Results

• Noise Floor Reduction: The active stabilization resulted in a drastic decrease in the DFS noise floor. The paper reports a greater than twofold reduction in the noise floor compared to the free-running laser.
• Detection Capability:
  • The system successfully detected a 120 Hz mechanical excitation at 101 km from the start of the fiber.
  • The signal emerged 10 dB above the noise floor in spectral analysis.
  • The system maintained a 1-meter spatial resolution and a 382 Hz mechanical bandwidth.
• Simulation vs. Reality: The experimental results confirmed the simulation predictions, validating that the specific low-frequency noise band identified in the model was indeed the primary culprit for sensitivity loss.

5. Significance and Impact

• Infrastructure Reuse: This work proves that existing telecom fiber spans (up to 100 km) can be repurposed for high-sensitivity environmental sensing (vibration, intrusion, structural health) without the need for expensive optical amplifiers or dedicated sensing fibers.
• Engineering Rules: The study provides concrete engineering guidelines for designing DFS systems, emphasizing that laser selection and stabilization must target specific low-frequency bands defined by the system's time parameters ($T_r$ and $T_p$).
• Scalability: By eliminating the need for Raman amplification and using standard telecom components (with active stabilization), this approach offers a cost-effective and scalable solution for large-scale distributed sensing networks.