Deliverable D3.2: Initial design of NI solutions and services exploiting zero-trust exposure and compute continuum layers

This document provides a comprehensive overview of network intelligence (NI) solutions for enhancing mobile network security, efficiency, and reliability. It emphasizes the role of machine learning, artificial intelligence, and advanced analytics in building the technology that supports ORIGAMI’s novel architectural components (namely, the Global Service Based Architecture, the Zero-Trust Layer and the Compute Continuum Layer). We structure this report around three key network domains: the Radio Access Network (RAN), the Transport Network (TN), and the Core Network (CN). We detail the design of our Network Intelligence (NI) solutions across RAN, TN, and CN domains, addressing the specific Barriers we identified in WP2. RAN solutions include task offloading, xApp conflict mitigation, and ML for energy efficiency across 11 use-cases. TN focuses on embedding ML in hardware for scalable user-plane intelligence, with 2 use-cases. CN presents 7 solutions for operational optimisation, security, and flexible operator models. We finally synthesize lessons learned from this work to further refine use-case definitions and support ORIGAMI’s architectural evolution.