Program

ENS 2026 PROGRAM

Program at a glance

Monday, June 29th, 2026

Room: Auditorium 2

09:00 - 09:10: Opening session

09:10 - 09:55: Technical session 1: AI-Driven Edge to Cloud Network Security and Optimization

09:55 - 10:30: Keynote 1: The Importance of Being EDGE

10:30 - 11:00: Coffee Break  

11:00 - 11:45: Technical session 2:  Emerging Techniques for Edge Orchestration and Edge Intelligence

11:45 - 12:20: Keynote 2: Edge-Defined Networking

12:20 - 12:30: Closing session

Detailed program

Monday, June 29 9:00 - 9:10 (Europe/Berlin)

Room: Auditorium 2

Opening session - Chair: Sebastian Troia

Monday, June 29 9:10 - 9:55 (Europe/Berlin)

Technical session 1: AI-Driven Edge to Cloud Network Security and Optimization
Chair: Sebastian Troia

Real-Time Intrusion Detection for IoMT with In-Network Inference on SmartNICs 

Aristide Tanyi-Jong Akem (University of Southampton, United Kingdom (Great Britain)); Noa Zilberman (University of Oxford, United Kingdom (Great Britain))

MERGE: Multi-Scenario Embedding for Robust Generalization in Network Anomaly Detection 

Christian Colella, Cristiano Serra, Alessio Sacco and Guido Marchetto (Politecnico di Torino, Italy)

Migration-driven Energy Consumption Optimisation atop Cloud Continuum 

Slawomir Kuklinski (Warsaw University of Technology, Poland); Robert Kołakowski (Orange Polska, Poland & Warsaw University of Technology, Poland); Bartłomiej Piotr Mastej (Warsaw University of Technology, Poland)

Monday, June 29 9:55 - 10:30 (Europe/Berlin)

Keynote 1 - Chair: Christian Grasso

Speaker: Guido Maier, Politecnico di Milano, Italy

Title: The Importance of Being EDGE

Abstract: The edge is nowadays regarded as a significant segment of the ICT infrastructure. We believe its importance is going to grow in the future, even if we are in an era dominated by the huge Cloud datacenters needed to support AI. The key to unleashing the potential of the edge is bringing the intelligence there, but this requires the deployment of suitable, robust, and affordable networking infrastructures. In traditional IoT systems, the edge is mainly used for the collection of data from the sensors on the far-edge towards the Cloud, where raw data is processed. Thus, the edge is simply a means to transfer bits. This works well for some applications, especially when sensors are simple, but it suffers from limitations when an added value is required from the network. The large difficulties 5G met to keep its promises of fostering new business in the IoT are, in our opinion, also related to these limitations.  On the other hand, the Cloud is gaining more and more relevance with the upsurge of massive AI, and in particular, the large-model AI, which requires huge hyperscaler datacenters to be trained. Systems entirely relying on LLMs are effective, but in some cases may raise issues: the user has no control over model development, sending data to the Cloud represents a loss of data ownership, and the processes involved require an enormous quantity of energy. Bringing the intelligence to the edge is the alternative solution with respect to completely relying on massive AI and relegating the edge to a mere bit-transportation function. It has several technical pros, such as making reactions faster due to reduced latency. But the most important advantage, in our opinion, is to allow the development of evolved services, based on AI model sharing between the communities and the possibility of keeping raw data at the far edge. It also provides an interesting opportunity for network operators and service providers to add value to their ICT assets.  This solution is particularly relevant for several verticals based on context awareness and territorial monitoring. In this keynote speech, we would like to share our research experience about the design and testing of distributed network/IT infrastructures to support an intelligent edge/far-edge. The use cases we will present are related to a recent and an ongoing research project: we hope these examples can stimulate discussion and new ideas about distributed, affordable edge architectures.

Monday, June 29 11:00 - 11:45 (Europe/Berlin) 

Technical session 2: Emerging Techniques for Edge Orchestration and Edge Intelligence
Chair: Alessio Sacco

Orchestrating Services with QoS Assurance at the Edge of Virtualized 5G Networks 

Gaetano Francesco Pittalà, Gianluca Davoli and Walter Cerroni (University of Bologna, Italy); Piotr Boryło (AGH University of Science and Technology, Poland)

Lightweight vs. Traditional Orchestration: A Performance Study of WebAssembly and Containers 

Nuutti Verneri Hakala (Ericsson Software Technology, Finland); Kashif Nizam Khan (Ericsson Software Technology & Aalto University, Finland)

Closing the Loop: Network-Aware Data Generation and Edge Learning in UAV-Based FANETs

Andrea Caruso, Christian Grasso and Giovanni Schembra (University of Catania, Italy)

Monday, June 29 11:45 - 12:20 (Europe/Berlin)

Keynote 2 - Chair: Gianluca Davoli

Speaker: Flavio Esposito, Saint Louis University, USA

Title: Edge-Defined Networking

Abstract: The notion of Software-Defined Networking has transformed how we design, operate, and reason about modern communication systems; indeed, it is the foundation upon which communities such as IEEE NetSoft were built. More recently, edge computing has evolved from a simple deployment model into a broader architectural paradigm that paved the way to increasingly intelligent, distributed, and latency-sensitive applications. In this talk, I will discuss recent efforts toward what I call ``Edge-Defined Networking": a paradigm where intelligence, programmability, and adaptation move closer to where data is generated, decisions are made, and actions are executed. In particular, I will explore how advances in programmable networking, distributed AI, and Cyber-Physical Systems are reshaping the role of the network edge from a passive execution environment into an active, adaptive, and intelligent control substrate. I will present examples of edge orchestration mechanisms spanning learning-based congestion control, O-RAN softwarization, and real-world edge deployments for smart sensing applications. I will then argue that future edge infrastructures will not simply host applications, but will increasingly participate in decision making through tight integration between communication, computation, and learning. This convergence raises important challenges in scalability, trustworthiness, sustainability, and real-time orchestration, while also opening opportunities for autonomous and self-adaptive networked systems.

Monday, June 29 12:20 - 12:30 (Europe/Berlin)

Closing session - Chair: Sebastian Troia