Convergence: Cognitive Intent Driven 5G Radio Access Network Slice Assurance

5G advanced network slicing use cases have gained momentum with the ability to provide differentiated Quality of Service (QoS) to various services (voice, mobile broadband, gaming, eXtended Reality, enterprise). However, it is essential to implement adaptive and scalable Network Slice Assurance to ensure differentiated service performance with varying traffic and dynamic requirements. The intent-driven network management paradigm aims at building autonomous systems with minimal expert-driven policies, which may be applied to slice assurance. In this paper, we propose Convergence, a cognitive intent driven framework for 5G Radio Access Network (RAN) slice assurance. Via the use of intelligent agents registered during the prediction, assurance, evaluation and actuation phases, emerging issues in slice assurance may be handled autonomously. In particular, Artificial Intelligence (AI) planning agents are exploited to generate partition over-use, partition top-up or partition ramp-down actions dynamically. The slice assurance framework is demonstrated over a real-world operator use case with multiple slices and intent specifications.