There has been substantial progress towards the ultimate goal, as highlighted by Marco Fiore, Associate Research Professor at IMDEA Networks and coordinator of the DAEMON project: “We have designed a number of original NI algorithms that have significantly advanced the state of the art in many practical networking functionalities, including virtualized Radio Access Network (vRAN) scheduling, control, and orchestration; Virtual Network Function (VNF) location and control; anomaly detection, prediction, and response; edge and slice resource orchestration; forward allocation of core network resources; and reconfigurable intelligent surface (RIS) control.”.

“The project has the potential to introduce changes to mobile network operations that will greatly benefit not only operators but also customers, who will enjoy better service with less delay, higher performance, and a more resilient infrastructure. Furthermore, part of the effort is dedicated to limiting the energy consumption of network infrastructure, thus reducing the footprint of mobile communications, which benefits society in these challenging times of global warming,” explains Dr. Fiore.
This European project, funded under the EU’s H2020-ICT-2020-2 call on Information and Communication Technologies, involves 12 partners across Europe, including major industrial players in the telecommunications ecosystem, innovative SMEs, and academic institutions. Dr. Fiore emphasizes that “it is precisely thanks to the close cooperation between industry and academia that the project is on track to demonstrate the viability of NI's algorithms in realistic environments, including 12 experimental test benches, 4 simulators, and data-driven testing with 14 different measurement datasets.”

The consortium also proposes an initial design for a new NI plane that complements the legacy user, control, and management planes that make up mobile network architectures. “The research conducted to date is definitely groundbreaking. It addresses many open problems in the automation of mobile network infrastructure operation and proposes innovative solutions based on a novel design of AI models that adapt to network environments or practical network architecture upgrades that improve its ability to adapt to self-management functionalities,” the project coordinator emphasizes.

Efficiency and Sustainability:
These are some representative examples of the technical challenges addressed by the project:
• Traditional AI algorithms make decisions to optimize a given performance metric, which in the network context is provided by a human team (e.g., network managers or systems engineers). However, in practice, the network operator does not always know in advance the performance metric that needs to be optimized through proactive network management actions. To address this challenge, DAEMON employs an innovative NI design based on so-called loss meta-learning. This is an original concept that allows the AI ​​algorithm to learn how its decisions affect the performance metric and how to optimize those decisions to maximize the performance metric in highly complex management tasks.
• DAEMON is developing AI-based solutions to achieve reliability in virtualized radio access networks hosted by highly shared resources, which reduce costs but cause fluctuations in computing capacity and thus pose a risk to reliability. The project partners are addressing the challenge by combining their expertise in the operation of a distributed unit (DU) and its radio processing operations with the predictive power of machine learning models.
The project is implementing tailored AI solutions for anomaly detection in IoT platforms that rely on international mobile roaming to connect IoT devices across various verticals (e.g., connected elevators, fleet tracking, smart meters, etc.). Given the complexity of the ecosystem supporting this global operating model favored by IoT verticals, AI algorithms are crucial for triggering alarms when IoT devices experience various anomalies.
The excellent progress of the research is evidenced by the high scientific impact of the project's results, which have been presented at all major peer-reviewed international networking conferences, such as ACM SIGCOMM, ACM MobiCom, IEEE INFOCOM, and ACM MobiSys, with over 49 papers published by the project to date.
Technology transfer is also a strength of DAEMON, as its consortium has currently filed three patents associated with solutions developed in the project, and more than 25 contributions to standardization bodies such as O-RAN, ETSI, and 3GPP. One of the innovations developed in the project by Telefónica, one of DAEMON's partners, was reviewed and subsequently accepted under the European Commission's Innovation Radar, a wide-ranging communication initiative supported by 23 EU Member States that aims to promote the latest technological and scientific advances made by European innovators.

Source: IMDEA Networks Institute