The 20th Anniversary Conference on Network and Service Management (CNSM) 2024 took place in Prague

Artificial intelligence and machine learning are spreading among most areas of academic, business, and personal lives of people around the world. Even networks and their security and privacy are not an exception and these topics open new research challenges. Therefore, over 150 researchers from 26 countries met in Prague for the 20th year of the Conference of Network and Service Management to discuss recent scientific breakthroughs and progress.

Benefits and pitfalls of artificial intelligence

Many technologies are no longer just a domain of IT experts, and we all observe the boom of large language models (LLM) that are catching attention globally. However, experts warn about various pitfalls such as hallucinating (when a model constructs false facts) or lack of proper reasoning. Dr. Noura Limam (from university of Waterloo in Canada) stated during the panel discussion of the Workshop on Analytics for Service and Application Management (AnServApp): “If I needed some advice to save someone’s life, I would definitely not trust ChatGPT.” However, the experts agree popular language models are amazing technology with big potential for the future, e.g., for generating various outputs in machine readable format, or as an interface for users in their native language, as it was discussed by distinguished experts prof. Luciano Paschoal Gaspary, dr. Carol Fung, prof. Gabi Dreo Rodosek, and Ing. Jiří Rott.

The participants and speakers focused on many topics like improving properties of machine learning applied on various domains especially network management and security. The goals involve explainability of models, evaluation and high-quality datasets. Data drifts in network traffic are considered a serious problem preventing ML-based classifiers and detectors to be deployed in practice. It remains a long-term enormous challenge to invent a proper handling of such drifts.

The future of network management and connectivity with artificial intelligence

Contrary, dr. Sebastián García, the keynote speaker from StratosphereIPS lab at Czech Technical University in Prague, revealed many exciting opportunities for use of AI in security and active defense. The closing discussion by the distinguished panel loosely followed with thoughts regarding the maturity of AI and LLM. According to prof. Gabi Dreo Rodosek, we need to face many challenges such as zero trust, zero touch in the future networks based on 6G and further. Automation and self-healing seem to be essential to build the future technologies working altogether. Cisco and its speakers showed an exciting state and future of the AI-powered products they are working on, whereas we need to consider both scopes “AI for networks” and “networks (infrastructures) for AI”.

Thanks to the organizers for the jubilee CNSM2024 in Prague

CNSM2024 was hosted in Prague and organized by joint effort of the team from CESNET and the Faculty of Information Technology, Czech Technical University in Prague. “It is a great honor and also a proof that the Czech Republic has already established itself in the scientific community that governs this annual conference. I am very proud of my team, which allowed us to host this prestigious event, and stand up besides the organizing countries of the previous CNMS events. Especially, I want to thank dr. Karel Hynek, my former Ph.D. student, who did an amazing job all over the whole organization.” commented dr. Tomas Čejka, the leader of the research team and local organizing committee co-chair.