Evolution of Networking Systems Software towards 6G
Thursday, 5 June 2025, 11:00 – 12:30, room 1.F
Session Chair:
- Nakjung Choi (Nokia Bell Labs, US)
- Mehdi Bennis (Univ. of Oulu, FI)
- Thierry Coupaye (Orange, FR)
Mobile networks are evolving into comprehensive software systems, where both operational functions and management capabilities are executed through a collection of SW modules that interoperate via a variety of APIs. This transformation is essential as future networks will not only converge fixed and mobile networks but will also, in the long term, integrate non-terrestrial, public, and non-public networks. Thus, as networks evolve toward 6G, the transformation of network systems software will be critical to realizing the envisioned 6G KPIs and use cases.
At the same time, the implications of softwarization and the adoption of emerging trends such as AI-native networking extend beyond mere implementation and network deployment aspects. By leveraging these technologies, telecommunication networks are not only becoming more flexible, self-organizing, resilient, and autonomous but are also required to do so. They integrate AI models of their own capabilities to expose capability-driven semantic interfaces, thereby enhancing network consumability and facilitating seamless integration with enterprise solutions and future applications such as the industrial metaverse. These networks inherently utilize AI for functions and autonomous management, dissolving traditional network boundaries by extending network capabilities through applications that become integral components of the network system. Additionally, they create network digital twins that interact with the digital twins of devices, applications, workflows, or entire production systems.
To that end, this workshop aims to provide a forum for researchers and professionals across academia, government, and industries, to exchange ideas, present new results, and provide future visions concerning the evolution of networking systems software towards 6G. Moreover, this workshop will focus on the conference track ‘NET – Network Softwarization’ but will also touch aspects of ‘AI4C – AI/ML Solutions for Communications’ and ‘NVS – Next-generation Visions and Sustainability’, by providing an overview on the 6G research from major academic institutions, network equipment vendors, service providers, and policy makers.
Programme
11:00 – 11:05: Welcome to the special session, Nakjung Choi, Nokia Bell Labs & Mehdi Bennis, University of Oulu & Thierry Coupaye Orange
11:05 – 11:20: Talk by Markus Gruber, Nokia Bell Labs
Title of the talk:
Network Optimization for New Challenges
The internet was originally not designed for low latency applications, so new architectural approaches are required to accommodate them; at the same time the symbiosis of the network and applications will entail semantic/goal-oriented communication approaches in order to meet the specific needs of the user. Secondly, disaggregation, an observable trend during the last decade, will continue to play an important role in the future. In this context, the interoperability between various ecosystems will gain importance while the basic principles of interaction converge.
11:20 – 11:35: Talk by Ricard Vilalta, CTTC
Title of the talk:
Intent-based Connectivity Service Provisioning with GenAI and TeraFlowSDN
Intent-based connectivity is an emerging approach that addresses the complexities involved in provisioning, managing, and assuring network services. By translating high-level requirements described as intents into policies that guide network behavior, intent-based networking reduces manual intervention and provides more agile service lifecycle management. This presentation examines how Generative AI (GenAI) has been integrated with TeraFlowSDN, a microservices-based network automation platform, to streamline connectivity service provisioning. The talk will explore the functionality of TeraFlowSDN and the role of GenAI in interpreting, refining, and validating intent statements, including bandwidth and latency specifications. Instead of manually configuring devices or adjusting parameters, network operators can state desired outcomes, allowing automated mechanisms to allocate and optimize resources accordingly. Attendees will learn about practical methods to test, validate, and operate an intentbased system in real-world scenarios. Case studies will illustrate how TeraFlowSDN harnesses AIenabled insights to satisfy user demands effectively while maintaining flexibility in network operations.
11:35 – 11:50: Talk by Sebastian Robitzsch, InterDigital Europe Ltd
Title of the talk:
Standard-Compliant Deterministic Networking for Multi-Domain and Multi-Technology 6G Systems
Many service flows between client and server endpoints traverse multiple computer networks even for a private network deployment. The challenge of reducing the complexity of such services is researched on by the EC-funded project PREDICT-6G, with special focus on deterministic communications. The provisioning of service flows over a multi-domain multi-technology group of computer networks is a particular challenge if each domain comes with its own set of deterministic communication support. To address that, an overarching E2E control plane is being developed in the project and is described herein. As the proposed innovations require a mix of technologies from SDO, this paper assesses the proposition of PREDICT-6G against the ongoing work in 3GPP, IETF, ETSI and IEEE.
11:50 – 12:05: Talk by N. Ghafoori, Iquadrat Informatica S.L., J.Vardakas, Iquadrat Informatica S.L., K.Ramantas, Iquadrat Informatica S.L. & Ch.Verikoukis, ISI/ATH and University of Patras
Title of the talk:
Energy-Efficient Intra-Domain Network Slicing for Multi-Layer Orchestration in Intelligent-Driven Distributed 6G Networks: Learning Generic Assignment Skills with Unsupervised Reinforcement Learning
As the 6th Generation (6G) of wireless networks emerges to meet the escalating demands of future connectivity, it introduces a highly complex, dynamic, and intelligent network architecture. Addressing this complexity requires scalable, decentralized, and energy-efficient management strategies. This research focuses on intelligent orchestration and automation in the edge and central domains of 6G networks, particularly emphasizing Network Slicing (NS) and resource allocation. We propose a comprehensive system model incorporating Open Radio Access Network (ORAN) and Cell-Free massive MIMO (CF mMIMO), delivering end-to-end slices from users to the Centralized Unit (CU). Each slice comprises a set of resource blocks and communication links tailored to service-specific Quality of Service (QoS) requirements. To enable intelligent and generalized slice configuration, we deploy a two-phase approach. First, a pre-training phase uses unsupervised Reinforcement Learning (RL) to discover transferable assignment skills across edge domains, independent of domain specifications. Second, Multi-Objective Reinforcement Learning (MORL) is employed to manage dynamic slice assignments in the central domain, enabling a single agent to handle multiple, potentially conflicting service types such as FeMBB and ERLLLC efficiently. This approach enhances scalability and reduces system complexity without compromising performance. Numerical evaluations confirm that the proposed system satisfies key 6G KPIs, including user-experience data rate and execution latency, ensuring reliable and adaptive service delivery across diverse network conditions.
12:05 – 12:20: Talk by Dorottya Papp, Roland Nagy, Jozsef Sandor, Andras Gazdag, Tamas Holczer, Gergely Biczok & Levente Buttyan: Budapest University of Technology and Economics
Title of the talk:
A novel threat taxonomy for 5G/6G networks as distributed systems
Both telecom vendors and operators are changing their viewpoint on the applied trust model in 5G/6G networks. Owing to the heavily softwarized nature of network functions and an increasingly complex hardware/software supply chain, malicious actors may be able to compromise assets well beyond traditional perimeter-based defenses. In this talk, we present a novel, extended threat taxonomy that takes into account the above changes in the network environment. We also discuss further challenges in applying the taxonomy for attack impact analysis and hint at the potential of zero-trust security mechanisms in mitigating these novel risks.
12:20 – 12:30: Closing the special session, Nakjung Choi, Nokia Bell Labs & Mehdi Bennis, University of Oulu & Thierry Coupaye Orange
