In conjunction with
22nd International Conference on Network and Service Management
Alcalá de Henares (Madrid), Spain // 26 October, 2026
Session 1 (13:30 - 15:00): Resource Optimization and Cost-Resilience Trade-offs
13:30 - 13:35: Opening
13:35 - 14:15: Keynote - Ed Fernandez: The Trade-offs of AI Infrastructure: From Research Questions to Reproducible Results
14:15 - 14:30: Cost-Robustness tradeoffs in optical access networks; Authors: Camilo Anzola-Rojas; Carmen Mas-Machuca
14:30 - 14:45: Computationally Efficient Max-Min Fairness Allocation for Specialized Multi-Site Networks; Authors: Noah Mehling; Frank Loh; Tobias Hoßfeld
14:45 - 15:00: When Is Redundancy Worth Its Cost? Resource-Resilience Trade-offs in Service Function Chain Protection; Authors: Trond Vatten; Malin Raanes Ludvigsen; Henny Løfqvist; Poul E. Heegaard
15:00 - 15:30: Coffee Break
Session 2 (15:30 - 17:15): Security, Traffic Engineering, and Protocol Innovation
15:30 - 15:45: Adaptive SASE-Driven Cybersecurity for V2X Communications in 6G Networks: Architecture and Preliminary Results from the SafeRoute-6G Project; Authors: Teo Redondo; Miguel Sanchez-Cuarental; Isabel Bou-Bayona
15:45 - 16:00: Robustness of Centralised, Localised, and Vertical Federated Anomaly Detection Against Data Manipulation Attacks; Authors: Andreas Flatscher; Michael Somma; Branka Stojanovic
16:00 - 16:15: Ontology-Based Network Digital Twin for Automated Availability Assessment of 5G Networks; Authors: Ikram Temericht; Firas Maalej; Guillaume Fraysse; Fabrizio Granelli; Bertrand Decocq; Yiping Fang; Zhiguo Zeng; Anne Barros
16:15 - 16:30: SAGE: Traffic Engineering for Heterogeneous Shared Memory Pools; Authors: Manuel Vögele; Nicolas Hornek; Andreas Kirstaedter; Timo Hönig; Benedict Herzog
16:30 - 16:45: Explicit AQM-based Congestion Notification for Improved Adaptive Video Streaming; Authors: Frank Loh; Alan Weber; Andrea Pimpinella; Fabio Martignon; Tobias Hoßfeld
16:45 - 17:00: Sleep Tight, QUIC Fast: Energy-Efficient QUIC with DPDK; Authors: Mohsen Memarian; Andreas Kassler; Johannes Späth; Marcel Kempf; Stefan Lachnit; Johannes Zirngibl; Georg Carle
17:00 - 17:15: MiNERVA: A Mininet Network Emulator for Reproducible Validation of Data Center Network Architectures; Authors: Moritz Freund; Sebastian Rieger
Detailed information, how the workshop program embedds in the main CNSM program is available here: https://cnsm-conf.org/2026/program.html
Ed Fernandez is co-founder and partner at 352 Capital Partners and a board director at BigML, a machine-learning platform company.
An electrical and electronics engineering graduate of the Universidad de Alcalá, he has completed the Global Senior Management Program (GSMP), jointly offered by the University of Chicago Booth School of Business and IE Business School, and the Engineering Leadership Professional Program (ELPP) at the University of California, Berkeley.
He has held international leadership roles at Nokia and BlackBerry, taught artificial intelligence and machine learning at Northeastern University, and currently serves as a trustee of the Universidad Politécnica de Madrid.
His work spans applied AI, technology investment and institutional governance.
Drawing on recent AI-platform design work and developments in scientific AI, this keynote explores sovereignty, security, resilience and sustainability as interacting systems requirements. It connects advances in structured decision models and evaluator-driven discovery with emerging questions of scientific evidence, provenance and research governance.
How can a research group turn a complex scientific question into an executable, reproducible workflow?
The talk examines AI infrastructure as the orchestration of reasoning models, specialised decision systems, scientific simulators and verification tools.
A 6G-oriented network research scenario, complemented by an aircraft-engineering example, makes the trade-offs concrete: model and workload routing, the protection of confidential research data, distributed computation, experimental validation and energy use.
These examples illustrate how architectural choices shape both scientific outcomes and the institution’s ability to govern, reproduce and sustain the research process.
The keynote concludes with an evaluation framework for comparing architectures against reproducibility, service quality, operational resilience and whole-workflow resource consumption.
Frank Loh, University of Würzburg, Germany (frank.loh@uni-wuerzburg.de)
Jacopo Talpini, University of Applied Sciences and Arts of Southern Switzerland, Switzerland (jacopo.talpini@supsi.ch)
Christoph Lipps, German Research Center for Artificial Intelligence, Germany (christoph.lipps@dfki.de)
Carlos Natalino Da Silva, Chalmers University of Technology, Sweden (carlos.natalino@chalmers.se)
Benedict Herzog, Ruhr University Bochum, Germany (benedict.herzog@rub.de)
Achim Autenrieth, Adtran Networks SE
Omran Ayoub, University of Applied Sciences of Southern Switzerland
Davide Borsatti, University of Bologna
Hesham ElBakoury, USA
Doğanalp Ergenç, Technische Universität Berlin
Poul E. Heegaard, Norwegian University of Science and Technology
Nguyen Huu Thanh, Hanoi University of Science and Technology
Stanislav Lange, NTNU
Piotr Lechowicz, Chalmers University of Technology
Shi Li, VPIphotonics GmbH
Chiara Lombardo, University of Genoa
Charitha Madapatha, Chalmers University of Technology
Carmen Mas-Machuca, University of the Bundeswehr Munich
Paolo Monti, Chalmers University of Technology
Petr Munster, Brno University of Technology
Beny Nugraha, Chemnitz University of Technology
Timothy O'Farrell, University of Sheffield
Jan Pennekamp, RWTH Aachen University
Andrea Pimpinella, University of Bergamo
Julie Raulin, Adtran Networks SE
Alessandro E. C. Redondi, Politecnico di Milano
Marco Savi, University of Milano-Bicocca
Andreas Schmidt, Saarland University
Florian Schreiner, Fraunhofer Institute FOKUS
Carlo Segat, TU Berlin
Michael Seufert, University of Augsburg
Federico Tonini, CNIT
Ruizhi Yang, University of Bristol