Feature Story | 6-May-2024

DAEMON project paves the road to effective 6G networks

The European project ends with a gamut of innovations for the integration of AI in mobile network environments

IMDEA Networks Institute

DAEMON, a European project coordinated by IMDEA Networks Institute that started in January 2021, has just come to an end, achieving significant milestones in the advancement of mobile communication technology. The project has paved the road to effective 6G networks, with clear benefits to society at large in terms of faster and more ubiquitous mobile communication infrastructures that can support new services.

The primary objective of this project was to demonstrate how AI can be practically integrated into production-level mobile network architectures and help automate the management of beyond-5G and 6G systems. The results have been a success since “the AI-native architectural model proposed by DAEMON has been adopted by the 5G PPP and 6G IA Architecture Working Groups, from where it has the potential to be further pushed into main standardization bodies like 3GPP”, states Marco Fiore, IMDEA Networks Research Professor and the DAEMON Project Coordinator. He continues: “Also, the project has generated two pilots that are now deployed in the production network of Telefónica, five solutions that have been included in the European Commission Innovation Radar initiative, nine patents, and over 100 scientific papers. Together, all these constitute the significant legacy of DAEMON”.

IMDEA Networks researchers, particularly from the Network Data Science group, have contributed to the research activities on customizing loss functions for networking-tailored AI, introducing concepts that are highly innovative for the Machine Learning community as well, such as that of loss meta-learning. In addition, they have demonstrated how AI models can be directly embedded in programmable network hardware for the classification of traffic flows with delays in the order of tens of nanoseconds.


The project tackled several challenges including:

– Understanding the limits of AI for mobile networks: they carried out a systematic, critical analysis of which network management tasks can be appropriately solved with AI models rather than other techniques and provided a solid set of guidelines for the utilization of machine learning for next-generation mobile network automation.

-Designing network management solutions empowered by highly customized AI techniques: based on the guidelines above, they developed and implemented a range of algorithms adapted to the unique requirements and constraints of mobile network environments that advanced the state of the art in automation for a precise set of core network management tasks.

-Designing an end-to-end NI-native architecture: they outlined the very first AI-native mobile network architecture model that, stemming from current standardization trends, enables the coordination of the many and varied AI instances deployed in the network.

High scientific production and the future of research

The research carried out in DAEMON has definitely been groundbreaking in nature. The solutions produced have been implemented and demonstrated in relevant production-grade network environments. At the same time, the innovative and high-quality research output of the project is attested by the many (over 30) publications at top-tier conferences and journals including ACM SIGCOMM, ACM MobiCom, IEEE INFOCOM, ACM/IEEE Transactions on Networking, or IEEE Transactions on Mobile Computing, among others.

While the DAEMON project has concluded, the journey of research continues with ORIGAMI, a follow-up European project initiated in January 2024. The quest for mobile networks with lower latency, higher throughput, and greater reliability progresses, promising improvements in various sectors and societal activities.

DAEMON has left an indelible mark on the evolution of mobile networks, and the future looks even more promising with projects like ORIGAMI driving innovation forward.

Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.