As the project progresses, and our solutions become more mature, peer-reviewed posters published by project partners, as well as project related posters will be made available here.
Return to this page for updates and new posters describing our work!
The MARE project is developing a programmable, modular and secure Security Plane for future 6G ecosystems, addressing the growing cybersecurity challenges created by highly interconnected “networks of networks,” AI-driven infrastructures and open programmable interfaces. By combining AI-powered orchestration, proactive stress-testing through Network Digital Twins and a modular library of deployable security building blocks known as DOTs, MARE aims to deliver adaptive, explainable and trustworthy cybersecurity across the entire 6G continuum. The architecture will be validated through 11 Proof-of-Concepts covering critical areas such as O-RAN, cloud-native environments, AI-driven security and far-edge networks, ultimately contributing to European digital sovereignty and the development of secure, resilient next-generation mobile networks.
This poster was presented at ICT2026 by the MARE Project Coordinator Dr. Andreas Zalonis.
This poster presents a Closed Loop Intent Based Agent designed to enhance resilience in multi-UPF 5G Core networks through autonomous detection and recovery from “gray failures” — subtle network degradations that occur while links remain operational. Leveraging free5GC, Uplink Classifier (ULCL) functionality and a custom Dynamic Steering API, the system continuously monitors network conditions and dynamically reroutes traffic between User Plane Functions (UPFs) without interrupting active services. Experimental evaluation using high-definition video streaming demonstrated that the agent can restore degraded services from poor quality levels (MOS 2.0) to good quality (MOS 3.5) in under 10 seconds, highlighting the feasibility of self-healing, intent-driven orchestration for future softwarized 5G and beyond networks.
This poster was presented at ICT2026 by Costas Gkinos.
This poster presents the EU project MARE’s approach to strengthening intent-based networking for 6G and beyond by integrating agentic AI and large language models into a security-focused architecture. It proposes a hierarchical, multi-agent, multi-domain, and multi-vendor framework in which an orchestrator agent translates high-level security intents into executable low-level actions, coordinates subordinate agents and tools, and can draw on external security knowledge bases such as MITRE ATT&CK, MITRE FiGHT, and NIST. The work also introduces an interactive intent-processing pipeline with human-in-the-loop support, designed to improve automation, explainability, and adaptability in complex network environments. A prototype implementation shows promising early results, indicating that even relatively small reasoning-capable language models can effectively support this architecture for secure and scalable future network management.
Co-funded by the European Union
MARE project has received funding from the Smart Networks and Services Joint Undertaking (SNS JU) under the European Union’s Horizon Europe research and innovation programme under Grant Agreement No 101191436
