6G-LEADER aims at evolving the PHY and RAN aspects of 6G communication networks by relying on the following pillars:
- ML-empowered PHY algorithms with predictive capabilities towards fully autonomous operation
- full-duplex transceivers employing novel sparse antenna arrays and advanced digital self-interference cancellation
- non-orthogonal and random multiple access schemes to accommodate mass connectivity demands of users and machines
- goal-oriented semantic communications
- an open and disaggregated RAN implementation with xApps integrating the advancements, achieved during the projects lifetime.
On this project, Eviden -an Atos business- will be the Task Leader of T6.2. We also will lead the development of AI/ML driven near real-time (xApps) and real-time (dApps) RAN control applications, whose performance will be evaluated and validated in two 6G LEADER Proofs of Concept (POCs).
The first POC - AI/ML Trainable 6G RIC Conflict Manager: Energy Efficiency & Critical Services - will be used to test the 6G-Leader Conflict Manager framework with Eviden's applications; the second POC - Wireless for AI based on AirComp and empowered by semantically-aware dApps/xApps - will be used by Eviden to integrate ML-driven resources and task allocation applications for AI services with different latency requirements, for a smarter network.
