Improving the fiber communication infrastructure with theoretical developments
Astrid Barreiro Berrio defended her PhD thesis at the Department of Electrical Engineering on January 21st.
Modern communication depends on high-speed, reliable networks, and optical fiber technology forms the backbone of this system. Whether it鈥檚 connecting cities, linking wireless base stations, or enabling long-distance communication, optical fibers ensure the rapid and efficient transfer of information. To keep up with growing digital demands, scientists are continuously working to optimize fiber-optic networks. In her PhD research, Astrid Barreiro Berrio tackled key challenges in fiber modeling, introducing innovative approaches to improve the first-order regular perturbation approximation (FRP) model鈥檚 accuracy and efficiency. Her work enhances fiber-optic modeling, making it more practical and scalable for future communication networks.
A major challenge in fiber-optic communication is understanding how signals travel through optical fibers. One key mathematical model, the Manakov equation, does not have a straightforward solution, making it difficult to fully predict and optimize fiber performance. To address this, researchers use the FRP technique, which is useful for describing fiber behavior but becomes inaccurate and complex when dealing with high-power signals. Barreiro Berrio introduced several innovations to overcome these limitations and improve the FRP model.
Main contributions
Barreiro Berrio developed an optimization method using gradient descent to refine key parameters, called kernels, within the FRP model. This improves accuracy while reducing complexity, making the model more applicable to real-world fiber networks.
She also introduced a geometric constraint methodology that identifies and restricts the set of FRP kernels to only the most significant ones. This simplifies calculations and enhances efficiency without sacrificing accuracy.
Her research compared different ways to simplify the FRP model, identifying magnitude-based pruning as the best approach to balance performance and complexity.
Finally, she refined the FRP algebraic expression to account for phase recovery at the receiver, broadening the range of scenarios where the model can be effectively used and leading to better signal reconstruction.
More efficient and accurate predictions
These advancements in optical fiber modeling have significant implications for telecommunications. By improving the FRP model, Barreiro Berrio鈥檚 work enables more efficient and accurate predictions of fiber behavior, reducing reliance on costly numerical simulations. This means faster, more reliable networks capable of handling the ever-growing data demands of internet users worldwide.
Ultimately, these innovations bring us closer to a more advanced, scalable, and efficient optical communication infrastructure, ensuring that the internet continues to evolve and support the needs of the modern world.
Title of PhD thesis: . Promotor: Prof. Alex Alvarado. Co-promotor: Dr. Gabriele Liga.