D5.2 Preliminary integration for AI-based E2E network slicing control and MANO

This document reports the up-to-date progress towards achieving smart E2E network slicing. It focuses on the prototyping of enabling technologies for E2E network slicing across network segments, their intent-based Management and Orchestration, and AI-based RAN resource allocation enabled with the Cognitive Plane’s Multi-Agent Deep Reinforcement Learning platform. Moreover, it covers preliminary integration work between different network slicing enablers over the E2E data path, with the MANO plane, and with the MA-DRL platform. Together, these technological enablers, algorithmic advances and initial integration work lay down the foundation for implementing the targeting E2E slicing system.