This document reports the design and prototype implementation of the foundation components that enable SliceNet Quality of Experience (QoE)-aware slice management. These enable the embodiment of the SliceNet Cognition Plane architecture described in deliverable D2.4. SliceNet QoE-aware slice management combines the established MAPE (Monitoring, Analysis, Planning, and Execution) autonomic control loop with state-of-the-art data-driven management and AIOPS (Artificial Intelligence for IT Operations); enabling intelligent, adaptive end-to-end (E2E) 5G slice management with respect to the use-cases (UCs) defined within the architecture of SliceNet. The components included in this report provide Proof-of-Concept (PoC) implementations covering the entire Cognition Plane; in particular, the required analytic methods, the machine-learning (ML) pipeline, QoE optimization, and vertical-informed QoE Actuators.
10.18153/SLIC-761913-D5_5 |
Rui Pedro, Guilherme Cardoso, Pedro Neves, Nuno Henriques, Xenofon Vasilakos, Nasim Ferdosian, Dean Lorenz, Kenneth Nagin, Marouane Mechteri, Yosra Ben Slimen, Albert Pagès, Fernando Agraz, Salvatore Spadaro, Rafael Montero, Antonio Matencio Escolar, Enrique Chirivella Perez, Jose M. Alcaraz Calero, Qi Wang, Ricardo Marco Alaez, Zeeshan Pervez
English
Network Slicing, AIOps, Network Slice, 5G, SDN, Cognitive management
application/pdf