State-of-Polarization Sensing with Coherent Transponders in Live Submarine Links: Noise Analysis, Processing Improvements and Practical Implications
This paper characterizes the noise limitations of state-of-polarization sensing in live submarine cables and demonstrates that advanced signal processing techniques can overcome estimation noise to enable robust, distributed earthquake detection across global submarine networks.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
Beneath the surface of the world's oceans lies a vast, silent network of glass threads that carry the internet, connecting continents with light. For decades, these submarine cables have been viewed strictly as communication arteries, their primary purpose to transmit data without interruption. However, the very physics that allows light to travel through these fibers also makes them sensitive to the environment around them. When the ocean floor shifts during an earthquake or when waves crash against the seabed, the glass fibers stretch and compress ever so slightly. This physical stress changes the way light travels through the cable, specifically altering the orientation of the light's waves, a property known as the state of polarization. While this change is usually a nuisance that engineers must correct to keep data flowing, a new perspective suggests that these corrections are actually a treasure trove of information. By listening to how the light twists and turns, we might be able to turn the global internet into a massive, continuous sensor for the Earth itself.
The challenge, however, has always been separating the signal from the noise. The light traveling through these cables is not perfectly steady; it is buffeted by the amplifiers that boost its strength over thousands of miles and the electronic equipment that reads it at the other end. These sources create a background hum that can easily drown out the subtle tremors of a distant earthquake. A team of researchers, working with equipment from Nokia and data from Sparkle's submarine cables, set out to understand exactly what this noise looks like and how to filter it out. They wanted to know if the existing technology, which is already installed on the ocean floor, could be tuned to detect earthquakes with enough clarity to be useful, or if the interference was simply too strong to overcome.
To answer this, the team analyzed data from ten active transponders on five different submarine cable routes in the Mediterranean Sea. These cables stretch from 250 to 2,000 kilometers, crossing diverse underwater landscapes. The researchers looked at the raw data that the equipment generates to keep the connection stable: a constant stream of measurements tracking how the light's polarization rotates as it travels. They discovered that the noise in these measurements follows a very specific pattern. At very low frequencies, the noise behaves like a slow, natural drift caused by the fiber itself changing shape over time due to temperature and pressure. But at higher frequencies, which are the most important for detecting earthquakes, the noise is flat and constant. This flat noise comes from the equipment itself—the electronic components and the digital processing inside the receiver. It is a static wall of interference that sits on top of the real seismic signals.
The researchers found that this flat noise is the main barrier to seeing earthquakes clearly. In the frequency range where most earthquake signals appear, between 0.1 and 1 hertz, the equipment's own noise is often louder than the tremors from the ground. This means that simply having the cable is not enough; the way the data is read and processed matters immensely. The team tested several methods to lower this noise floor. They found that taking measurements much faster than the standard rate significantly reduced the interference. By increasing the sampling speed from 2 times per second to 100 times per second, they could lower the noise floor by a factor of fifty, making the underlying signals much clearer. They also discovered that by combining data from multiple channels of light traveling through the same cable, they could cancel out random errors, further sharpening the view.
Beyond just reducing noise, the team developed new ways to process the data to avoid mathematical errors that can occur when tracking continuous rotation. They used a method based on differential tracking, which looks at the change between one moment and the next rather than trying to calculate an absolute position. This approach not only reduced the computational load by more than half but also prevented the data from becoming garbled during long periods of monitoring. When they applied these improved techniques to real-world data, the results were striking. The researchers successfully identified seventeen different earthquakes, ranging from moderate events to massive quakes, occurring between 65 and 750 kilometers away from the cables. The data showed a clear relationship: larger earthquakes produced stronger signals, and signals grew weaker as the distance from the cable increased.
The study confirms that the existing infrastructure of the global internet can serve as a powerful tool for geophysical monitoring, provided the equipment is configured correctly. The researchers showed that the sensitivity of these cables is not fixed; it depends heavily on the quality of the transmission and the speed at which the data is collected. With the right processing strategies, the system could potentially detect earthquakes as small as magnitude 3.6, a threshold that would be invisible with current standard settings. This does not replace the need for dedicated seismometers, but it offers a way to monitor vast stretches of the ocean floor where placing instruments is difficult or impossible. By turning the noise of the network into a signal, the researchers have demonstrated that the cables carrying our digital lives are also listening to the planet, waiting for us to learn how to hear them.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.