Skip to main content
European Commission logo
English English
CORDIS - EU research results
CORDIS
CORDIS Web 30th anniversary CORDIS Web 30th anniversary

Fundamentals of the Nonlinear Optical Channel

Periodic Reporting for period 4 - FUN-NOTCH (Fundamentals of the Nonlinear Optical Channel)

Reporting period: 2022-07-01 to 2023-06-30

Fiber optics can carry very large amounts of data at very high speeds. Indeed, nearly all the global Internet traffic is transported by optical fibers. However, the current traffic demands growing by a factor larger than 10 every decade, and thus, it is unclear if these fibers can continue carrying all the data across the globe. Maintaining the transport of digital information across the world at high speeds and reliably is of vital importance for society, specially now that more and more services heavily depend on digital systems. To meet capacity demands, fiber optical communication systems are being operated at high powers, bringing these systems into the nonlinear regime of the optical fibers. This regime is not well understood from an information and communication theory point of view. The overall objective of this project is to unveil the fundamental limits behind data transmission of fiber optics in the highly nonlinear regime.
The work was divided into 5 work packages (WPs). The research carried out in the project focused on mathematical modeling (WP1 and WP2) as well as to design new transceiver structures (WP4). Multiple experimental validations were performed (WP5). Capacity results for soliton-based communications and multiuser information-theoretic approaches have been obtained in WP3.

The specific results obtained in the project are the following:

In WP1 and WP2, different mathematical models were developed. These include data-driven models based on first-order perturbation theory (WP1 and WP2) and nonlinear interference models for dual-polarization (four-dimensional) formats (WP1).
In WP3, maximum transmission rates (or bounds thereof) of soliton-based communication systems have been performed. Multiuser information theory was also used to develope inner and outer capacity bounds. These results were based on the models developed in WP2.
In WP4, multiple contributions to the areas of channel coding and signal shaping were made. We pioneered the use of low-complexity (finite-blocklength) probabilstic shaping based on the idea of enumerative sphere shaping. We also developed new algorithms that combine soft and hard decisions from the optical channel and developed novel four-dimensional modulation formats. These algorithms and modulation formats have the potential to be used in next-generation ultra-high speed optical transponders.
In WP5, we experimentally validated multiple transceiver designs from WP4.

The project generated 35 journal papers, 11 Invited Talks/Conference papers, and 27 conference papers. The project also generated IP in the form of three patents. The results of the project were disseminated via presentation on workshops and conferences as well as through the website of the project that included all scientific publications in the project. We also used social media (twitter and linkedin) to engage with a broader audience.
All the work described above goes beyond the state of the art and is demonstrated by the fact that these works have been published in highly prestigious journals (Nature Communications, the IEEE Transactions on Communications, and the IEEE Journal of Lightwave Technology). The results in this project have also been presented at the most important conferences in the fiber optical research: The Optical Networking and Communication Conference & Exhibition and the European Conference on Optical Communication. Our work received two awards: the Best Paper Award at the 2018 Asia Communications and Photonics Conference and the Optical Student Paper Award at the Signal Processing in Photonic Communications (SPPCom 2022).
Sphere shaping vs other shaping algorithms
Capacity region inner and outer bounds
Normalized batch gradient descend algorithm
New channel model using perturbation on dispersion term