Optical fiber losses in quantum networks

Juan Moreno
6 min readOct 14, 2023

--

If you’re working as a network engineer in a decently-sized data center, you’ve probably dealt with fixing connectivity hiccups in your fiber optic links between racks. I recently had an eye-opening moment when we were setting up a quantum network and our link wasn’t behaving as expected, even though we were using top-notch fiber optic gear. The problem turned out to be a real head-scratcher. It was all about the power of the light signal we were sending from one rack to another. It was like a superhero with too much strength — it was overwhelming the receivers on the other end, causing errors and messing up the performance. The fun part was that we got to play detective and figure out what was going on. Our solution? We brought in this little gizmo called an optical attenuator. We attached it to both cable ends, adjusting the signal’s intensity to the perfect level. Not only did we solve the problem, but it also turned into a memorable learning experience.

Optical attenuators are like volume knobs for optical signals, allowing to reduce their power level. In certain high-power fiber optic applications, reducing the intensity of a signal can help mitigate non-linear effects, potentially optimizing its performance, which is rather useful to say the least in telecommunications, data centers, optical networks, and test and measurement setups. Furthermore, a good attenuator does the job without introducing significant signal distortion, ensuring integrity of the devices that are integrated in a fiber optic network. When we are setting up a short-span quantum network (up to a 100 km), a link encryptor is receiving key material from a Quantum Key Distribution (QKD) device, and sending these keys to a mirrored pair at the other side of the network (the jargon usually refers to them as “Alice” and “Bob”). In this type of setup, it is super important to avoid saturation of downstream components, such as receivers or transceivers.

Let me tell you a bit more about these little miracle-makers. An attenuator can be fixed or variable in terms of their attenuation level. Fixed optical attenuators give you a constant amount of attenuation, typically specified in decibels (dB), regardless of the signal power. Variable optical attenuators, on the other hand, offer tunable attenuation levels, allowing for fine-grained control of the optical power level. The latter are often used in applications where dynamic or on-the-fly adjustment of signal power is required, such as optical network testing, wavelength division multiplexing (WDM) systems, or in applications where signal power needs to be adjusted for varying link distances or environmental conditions.

Attenuation is a critical feature in quantum networks, which rely on the principles of quantum mechanics for information transfer and processing. In a quantum network, fiber optic cables transmit quantum information encoded in quantum states of light over long distances. Having high a quality link is extremely important, because optical fiber losses can significantly degrade signal quality, transmission distances, and ultimately limit the scalability and feasibility of quantum communication systems. Unfortunately, preventing losses is not always possible (or too expensive to implement when it is), so when doing a network qualification for a quantum project, you should know at the very least what types of fiber loss mechanisms are there. There are lot of details and nuance involved on characterizing fiber performance. I’ll focus here on categorizing fiber losses, and let the mechanisms to prevent them for a future discussion.

Source: https://community.fs.com/blog/understanding-loss-in-fiber-optic.html

Intrinsic losses

Intrinsic losses in optical fiber are essentially the downsides caused by the fiber’s inherent characteristics. It’s like when you’re trying to enjoy a sunny day, but suddenly, clouds roll in. In this scenario, our “clouds” are absorption and scattering. Absorption losses occur when the fiber material absorbs some of the incoming light, converting it into another form, such as heat. Imagine your sunny day becoming unexpectedly warm. Scattering losses happen when the fiber redirects the light in various directions. This optical deflection results in a weaker transmitted signal. Among these scattering phenomena, there’s a prominent one known as linear Rayleigh scattering. This is caused by tiny variations in the fiber’s structure, scattering the light in multiple directions, similar to how confetti flies around during festivities. Understanding these intrinsic losses is like determining why your sunny day can become cloudy or why some surprises don’t always work out. It’s all part of enhancing optical communication systems.

Losses due to contaminations

Contaminations in optical fibers represent disruptions similar to unwelcome interruptions in an orchestrated event. These disruptions, stemming from impurities or defects, jeopardize the clear passage of light through the fiber.

For instance, metal ions, often introduced during manufacturing, handling, or deployment, can scatter light, diminishing its efficacy. Originating primarily during the fused silica production process, these ions act as inadvertent barriers to the light’s journey.

Another notable contaminant is Hydroxyl radicals (OH). They present problems primarily through absorption. These radicals, resonating at frequencies akin to the infrared light wavelength, absorb the light with an unsettling efficiency. It’s as if a highly absorbent material is constantly ready to soak up the light’s energy.

Navigating these contaminations in optical fibers is a bit like troubleshooting glitches in our favorite gadgets. It’s all about understanding the hiccups, finding solutions, and ensuring our optical communication systems are running smoothly and at their best.

Induced fiber losses

Induced fiber losses are the unexpected glitches in the optical fiber journey, and they come from the outside world. These disruptions are like curveballs thrown into a smooth game.

These losses are caused by external factors that sneak their way into the manufacturing process. It’s a bit like when you’re painting a masterpiece, but suddenly, a gust of wind blows some dirt onto your canvas. These factors can range from hiccups in the fiber drawing process to all the bending and twisting that occurs during the coating and cabling phase.

Picture it this way: any dielectric waveguide, our fancy term for an optical fiber, prefers to keep things perfectly straight. But when it gets bent or curved beyond what it can handle (its minimum bend radius), things start to go off-script. It’s like trying to fold a piece of paper too many times — it creases, crinkles, and doesn’t behave as expected. In the case of the fiber, this results in the light losing its strength, and its signal becomes weaker. It’s like the movie suddenly becoming blurry and hard to follow.

So, induced fiber losses are like that sprinkle of chaos in an otherwise orderly world. Even in the realm of optical fibers, there’s always a bit of unpredictability keeping us on our toes.

In addition to using optical fibers that don’t lose much light and making sure we handle them carefully to avoid bending and scattering mishaps, one super effective way to keep our quantum networks going strong is to bring the so-called quantum repeaters. They’re the holy grail of the quantum world, helping us out over long distances. Repeaters are old pals in regular networks, but they’re all the rage in the quantum scene too. By getting these repeaters into the game and finding ways to correct errors, we’re making sure our quantum networks perform their best. That means we can send quantum signals over longer distances, make them super reliable, and even scale up the whole quantum communication and computing game. And that is like unlocking the secret powers of quantum technology. 🚀✨

--

--

Juan Moreno

Nomad. Addicted to jazz, books and coffee. The postings on this site are my own and don't necessarily represent the position of my employer.