AI-Enabled Network Slicing Orchestration for Scalable 6G Edge–Cloud Architectures

Abstract

The paper responds to the urgent requirement of dynamic and smart slicing in the sixth-generation (6G) wireless networks that are envisioned to provide ultra-reliable, low-latency and high-throughput communication. As the supporting service requirements start to increase (enhanced mobile broadband (eMBB) to ultra-reliable low-latency communication (URLLC) and massive machine-type communication (mMTC) ), the traditional static orchestration mechanisms are no longer sufficient to ensure scalability and the quality of service the support. In this respect, to fill this gap, we suggest an AI-enabled edge RAN slicing orchestration system built on edge-cloud 6G structures. The framework rests on machine learning (ML) to realize slice lifecycle automation, predictive-based resource allocation, and adaptive slice reconfiguration. It uses federated learning to enable decentralized intelligence and reinforcement learning to enable proactive adaptation using traffic dynamic and constraints of the service-level agreement (SLA). Our system-level design uses distributed AI agents at the edge, which is how they can execute important tasks at a low-latency level, and cloud-assisted policies can guarantee global coordination and long-term optimization. Real-as-tested simulation-based assessments of 6G real-time traffic patterns indicate that the presented approach would result in slice resource-efficiency improvements of 30 percent along with a 25 percent. These findings show the promise of AI-based orchestration in enabling 6G-based infrastructures that are scalable, with SLA guarantees, and service differentiated. The architecture is the stepping stone to robust and self-sufficient 6G networks that could meet the ability to adapt to the changing and heterogeneous user requirements across verticals industries.