Main achievements organized by the project's core technical objectives are:
* Objective 1: Evolving the Mobile System Architecture
Significant achievements: (i) Detailed design of the GSBA, which extends the 5G Service-Based Architecture to create a unified, multi-domain framework. This was advanced by proposing a functional split of the Radio Access Network's (RAN) Distributed Unit (DU) into control and user plane components, enhancing interoperability across traditionally siloed domains Ii) detailed design of the CCL , as a sophisticated abstraction layer to manage a wide array of heterogeneous computing resources, from standard CPUs and GPUs to quantum processors and (iii) Detailed design of the ZTL, aimed to modernize obsolete trust models (Barrier #6), is a security framework operating on a "never trust, always verify" principle.
* Objective 2: Designing Advanced Network Intelligence (NI)
Significant achievements: Developed a novel 6G architecture by designing a suite of advanced Network Intelligence solutions addressing critical, high-impact challenges. For example, in the RAN domain, the project designed and validated CloudRIC, an NI solution for sustainable vRANs (Barrier #1)). By coordinating heterogeneous computing resources, CloudRIC ensures 99.999% reliability for processing deadlines while demonstrating experimental gains of up to a 6x improvement in energy efficiency and a 40x improvement in cost efficiency. Another example for the Transport Network: the project addressed the under-utilization of programmable hardware (Barrier #4) by developing DUNE (Distributed User-plane INference).
* Objective 3: Enabling Global Service Deployment and Automation
Significant achievements: (i) For global operations and trust (Barriers #5 and #6), the GMNO-1 use case produced the design for NOMADIX, a nomadic User Plane Function (UPF) with separated control and data planes. This allows the data-forwarding component to be dynamically deployed in the cloud close to the end-user, delivering a local-like performance experience for roaming users. This, combined with the DICE protocol, provides a powerful technical foundation for a new generation of global operators. (ii) To tackle inadequate data representation (Barrier #7), the KR use case successfully leveraged Large Language Models (LLMs) to process and classify network incident tickets, automating the extraction of structured knowledge from unstructured text. To reduce control-plane overhead (Barrier #8), the NCAM use case analyzed the performance of a Service Communication Proxy (SCP) within a service mesh architecture. Experimental validation demonstrated that this model reduces signaling traffic from key network functions by up to 39% compared to traditional direct communication, confirming a path to a more scalable and efficient 6G core.
