Optical networks and artificial intelligence (AI) are increasingly driving each other’s development. While AI makes the operation of modern communication networks more efficient, these networks require high-performance fiber-optic infrastructures. Researchers at the Fraunhofer Heinrich Hertz Institute (HHI) are contributing their latest research findings to the landmark ETSI white paper “Strategies for AI Adoption in Fixed Networks: Challenges, Use Cases, and Future Directions,” thereby contributing to the international standardization of AI-supported networks.
The work reflects Fraunhofer HHI’s long‑standing expertise in AI‑enabled network automation, optical communication networks, and data‑driven network management. Researchers from Fraunhofer HHI contributed key content on machine learning based network automation, large language models (LLMs) for network operations, digital twins, and data governance for AI‑driven networking.
"We see that AI in fixed networks is no longer a purely academic topic – it is moving into real-world deployment. However, this transformation will be realized only through trusted, governed, and standardized integration of AI in fixed networks. That is precisely where our work comes in."
— Dr. Behnam Shariati
From the Lab to a Global Standard
The European Telecommunications Standards Institute (ETSI) develops technical standards for digital infrastructures worldwide. The white paper was initiated by Dr. Behnam Shariati at ETSI. He took on the editorial leadership and assembled a consortium of 14 organizations from industry, network operations, and research across four continents. The goal is to jointly establish technical foundations for the deployment of AI in fixed networks and to actively contribute European research perspectives to international standardization processes.
Contributing partners: Deutsche Telekom, Nokia, Telefónica, Huawei, China Telecom, MTN Nigeria, RISE (Sweden), UC3M (Spain), POST Luxembourg, Dacoso, Infosim, Lunet, Savantic, Waystream
Common Standards as a Requirement
Many of the technological foundations for AI-enabled networks are now considered ready for deployment. However, common standards are still lacking. Without them, there is a risk of fragmented solutions, increased operational risks, and missed economic opportunities within the European telecommunications market. At the same time, issues regarding data protection, transparency, and the explainability of AI decisions in critical infrastructures are gaining importance.
The ETSI white paper is available here.
Contributing authors: Angela Mitrovska, Johannes Fischer, Hussein Zaid, Mihail Balanici, Pooyan Safari, Ronald Freund (Fraunhofer HHI), along with experts from China Telecom, Dacoso, Deutsche Telekom, Huawei, Infosim, Lunet, MTN Nigeria, Nokia, POST Luxembourg, RISE, Savantic, Telefónica, UC3M, and Waystream.