This White Paper provides an overview of adopting AI in fixed networks, moving from individual siloed research activities and proofs of concept (PoC) to real-world deployments and delivering an AI-native, intent-driven, self-operating infrastructure. It explains how advances from statistical machine learning (ML) to deep learning (DL) and large language models (LLMs) are reshaping network planning, operations, assurance, security, and customer experience. It also briefly addresses “Networks for AI,” outlining the transport and data centre interconnect upgrades required to support AI workloads. The report consolidates lessons from proofs of concept and live deployments, identifies gaps hindering scale, and discusses potential actions for ETSI and the industry to standardise interfaces, data, governance, and assurance of AI systems in multi-vendor fixed network environments.

Accurate energy consumption forecasting is critical for optimizing energy usage, lowering operational costs, and encouraging sustainability in smart buildings. Machine learning (ML) has developed as an effective method for energy forecasting, using sensor data to anticipate consumption trends and increase efficiency. However, due to regulations such as GDPR and growing privacy concerns, sharing sensitive energy data with third parties is often prohibited, providing issues for traditional centralized ML techniques. Federated Learning (FL) provides a feasible alternative by allowing for decentralized model training across several buildings without explicitly exchanging raw data. This privacy-preserving strategy enables organizations to jointly train reliable models while retaining data sovereignty. Our experimental results demonstrate that by using the CU-BEMS dataset, both FL and centralized forecasting models perform similarly, with an R2 score of ≈ 87%. Furthermore, FL decreases bandwidth use by limiting data transfers, making it a scalable and economical energy management solution for smart buildings. These findings demonstrate FL's ability to ensure safe, data-driven decision-making for sustainable energy utilization

Communication networks are vital for society and network availability is therefore crucial. There is a huge potential in using network telemetry data and machine learning algorithms to proactively detect anomalies and remedy problems before they affect the customers. In practice, however, there are many steps on the way to get there. In this paper we present ongoing development work on efficient data collection pipelines, anomaly detection algorithms and analysis of traffic patterns and predictability.

Cooperative Intelligent Transport Systems (C-ITS) enable information to be shared wirelessly between vehicles and infrastructure in order to improve transport safety and efficiency. Delivering C-ITS services using existing cellular networks offers both financial and technological advantages, not least since these networks already offer many of the features needed by C-ITS, and since many vehicles on our roads are already connected to cellular networks. Still, C-ITS pose stringent requirements in terms of availability and latency on the underlying communication system; requirements that will be hard to meet for currently deployed 3G, LTE, and early-generation 5G systems. Through a series of experiments in the MONROE testbed (a cross-national, mobile broadband testbed), the present study demonstrates how cellular multi-access selection algorithms can provide close to 100% availability, and significantly reduce C-ITS transaction times. The study also proposes and evaluates a number of low-complexity, low-overhead single-access selection algorithms, and shows that it is possible to design such solutions so that they offer transaction times and availability levels that rival those of multi-access solutions.

This final report is the second of two reports. Both are the result of a Swedish project, the Swedish Energy Authority’s TENS project, which spanned from 2018-2021. Other reports from the project provide results from estimating emissions from traffic measurements as well as simulation studies. Like any European capital, Stockholm suffers from many problems related to its road network. The main factor is traffic jams, which are aggravated with difficult weather conditions in winter but also due to accidents, popular events and holidays. Therefore, this report provides the results from a data-driven approach to estimating traffic flow. This work aims at predicting and understanding the behavior of this network based on data collected at several places. More specifically, the goal is to predict and model the traffic flow i.e macroscopic information, on ground measurements (MCS), using floating microscopic (INRIX) data. We focus on estimating the fundamental traffic law relationships, the flow using time series and future directions. Methods and results are in the related work section.

ANIARA (https://www.celticnext.eu/project-ai-net) attempts to enhance edge architectures for smart manufacturing and cities. AI automation, orchestrated lightweight containers, and efficient power usage are key components of this three-year project. Edge infrastructure, virtualization, and containerization in future telecom systems enable new and more demanding use cases for telecom operators and industrial verticals. Increased service flexibility adds complexity that must be addressed with novel management and orchestration systems. To address this, ANIARA will provide en-ablers and solutions for services in the domains of smart cities and manufacturing deployed and operated at the network edge(s).

ANIARA (https://www.celticnext.eu/project-ai-net) attempts to enhance edge architectures for smart manufacturing and cities. AI automation, orchestrated lightweight containers, and efficient power usage are key components of this three-year project. Edge infrastructure, virtualization, and containerization in future telecom systems enable new and more demanding use cases for telecom operators and industrial verticals. Increased service flexibility adds complexity that must be addressed with novel management and orchestration systems. To address this, ANIARA will provide en-ablers and solutions for services in the domains of smart cities and manufacturing deployed and operated at the network edge(s). © 2021 Owner/Author.

We present the latency-aware multipath scheduler ZQTRTT that takes advantage of the multipath opportunities in information-centric networking. The goal of the scheduler is to use the (single) lowest latency path for transaction-oriented flows, and use multiple paths for bulk data flows. A new estimator called zero queue time ratio is used for scheduling over multiple paths. The objective is to distribute the flow over the paths so that the zero queue time ratio is equal on the paths, that is, so that each path is ‘pushed’ equally hard by the flow without creating unwanted queueing. We make an initial evaluation using simulation that shows that the scheduler meets our objectives.

Connected vehicles can make roads traffic safer andmore efficient, but require the mobile networks to handle timecriticalapplications. Using the MONROE mobile broadbandmeasurement testbed we conduct a multi-access measurementstudy on buses. The objective is to understand what networkperformance connected vehicles can expect in today’s mobilenetworks, in terms of transaction times and availability. The goalis also to understand to what extent access to several operatorsin parallel can improve communication performance.In our measurement experiments we repeatedly transfer warningmessages from moving buses to a stationary server. Wetriplicate the messages and always perform three transactionsin parallel over three different cellular operators. This creates adataset with which we can compare the operators in an objectiveway and with which we can study the potential for multi-access.In this paper we use the triple-access dataset to evaluate singleaccessselection strategies, where one operator is chosen for eachtransaction. We show that if we have access to three operatorsand for each transaction choose the operator with best accesstechnology and best signal quality then we can significantlyimprove availability and transaction times compared to theindividual operators. The median transaction time improves with6% compared to the best single operator and with 61% comparedto the worst single operator. The 90-percentile transaction timeimproves with 23% compared to the best single operator andwith 65% compared to the worst single operator.