22nd AIAI 2026, 16 - 19 July 2026, Chania, Crete, Greece

6G-PATH: Architecture Validation via Internal and Open-Call Trials

Corici Andreea, Zope Hemant, Gupta Peenal, Hocquel-Hans Martin, Sengupta Banhirup, Ferreira Luis, Corici Marius, Roß Leon, Lourenço Mario, Ellenberger Moritz, Blaeser Andreas, Cordeiro Luis, Fernandes João, Rosa Luis, Lahmann Nils, Strube-Lahmann Sandra

Abstract:

  This paper presents the 6G-PATH architecture and its validation through internal deployments and open-call trials in a healthcare scenario for chronic wound management. Starting from requirements derived with clinical partners and a survey of nursing personnel (N=146), we identify high availability, security, and seamless access to medical data as primary needs, while “futuristic” ultra‑low‑latency features rank lower. These findings drive the design of an end-to-end platform that combines a beyond‑5G core network with edge capabilities, a secure experimental testbed, and healthcare-specific enablers. On the network side, the platform integrates Open5GCore, local breakout at the edge, and mechanisms for policy enforcement \cite{3gpp_23501,3gpp_23502,etsi_mec_003}, Security as a Service, and reliability assurance over terrestrial and satellite backhaul. Healthcare applications are supported by a Kubernetes‑native \cite{kubernetes} Close Loop Coordinator for Artificial Intelligence (AI) pipeline orchestration across cloud–edge resources, QoS‑aware and interoperable bioprinter control, and PATHSHIELD, a 6G‑enabled security framework that uses federated learning for malicious device and attack detection and traffic influence for dynamic isolation and mitigation. A realistic testbed with on‑premises central and edge nodes and private 5G network on wheels enables trials under strict time and budget constraints. The validation confirms that the proposed architecture can meet clinical workflow needs while demonstrating how Beyond 5G (B5G) /6G capabilities—network slicing, Multi-access Edge Computing (MEC), federated learning for attack detection and traffic influence for its mitigation—can be concretely exploited in safety‑critical healthcare environments.  

*** Title, author list and abstract as submitted during Camera-Ready version delivery. Small changes that may have occurred during processing by Springer may not appear in this window.