The workshop on Realistic AI Usage in Networks addresses the challenges and constraints associated with implementationon and deployment of AI and ML methods in current and next generation networks (6G). Recent studies have unveiled that the actual deployment of AI/ML in networks offers only modest enhancements over classical means, while demanding excessive energy, data, time and other resources. Scholars recently started raising concerns related to difficulties in reproducing even those modest gains in a reliable manner. Finally, the non-deterministic nature of promised outcomes makes it hard to use AI-based methods in many situations.
In response to these justified doubts, this workshop aims to foster discussions among researchers to identify pragmatic approaches and potential solutions with/for AI/ML methods to fully leverage their potential in future networked environment such as 6G networks. The workshop will focus on key areas of concern, including model distribution, handling large models, real-time data acquisition, and decision-making processes. Participants will engage in sharing insighƞul research and experiences, seeking to lay the groundwork for a more responsible and sustainable integration of AI in network settings.
By promoting collaborative exchanges of knowledge, the workshop seeks to contribute to a deeper understanding of Realistic AI Usage in Networks, setting the stage for advancements that align AI technologies with the practical needs of users and operators in network environments.
Each session will have three speakers with 15min presentation and will be followed by 45min of discussion between the participants. Focus will be on the challenges we need to face ahead.
Objectives of the workshop, operations
Achievements and limitations
Where have we seen AI as really useful in telecom networks: areas of impact, tests, measures, trials highlighting those achievements and identifying limitations
Models for quantifying AI benefits. Costs of AI usage and training. Explainable and trustable AI. Privacy aspects
Network Data for AI
Which data to use and how? How to acquire, process and validate the data. Relevance of the data, and open data for e.g. 6G
Implications for future architectures, most notably for 6G. Computing and transmission costs, and architectural challenges. AI plataforms.
Speakers are being invited from CTTC, Deutsche Telekom, Huawei, Instituto de Telecomunicações, Nokia, Telefonica, Trinity College Dublin, University of Napoli, and Vodafone among others.
Organization is under the scope of Networld Europe (https://www.networldeurope.eu/) with the support of SNS OPS project (https://smart-networks.europa.eu/csa-s/#SNS-OPS), and the event will be adjacent to the Visions of Future Communication Summit (https://futurecomresearch.eu/).
If you want to participate or contribute, please contact the organization below.
Agenda will be closed by early October, based on the contributions received.