Search papers, labs, and topics across Lattice.
This paper introduces SubEdge, a novel subsystem designed for 6G networks that enables per-subscriber provisioning of integrated communication and computing resources, addressing the limitations of existing Net4AI architectures. By binding a Subscription Permanent Identifier (SUPI) to a computing context that includes inference containers and service entitlements, SubEdge facilitates seamless mobility-driven migration of compute instances and traffic-routing policies. Experimental results demonstrate a significant reduction in latency from 22.9 ms to 12.2 ms during mobility events, while maintaining high frame delivery rates and successful migration operations for multiple subscribers.
SubEdge slashes latency by nearly half during subscriber mobility, ensuring uninterrupted AI service delivery in 6G networks.
Beyond traditional connectivity, 6G is envisioned to transform mobile networks into a distributed fabric that provides native integrated communication, computing, and intelligence services. AI-native terminals (e.g., robots, autonomous vehicles, and smart glasses) require real-time inference from individualised, manufacturer-specific models that cannot be executed on-board nor shared across subscribers, making per-subscriber edge compute the necessary complement to per-subscriber connectivity. Existing Network for AI (Net4AI) architectures provision compute for application providers through shared deployments and do not address per-subscriber provisioning. This paper proposes SubEdge, a Net4AI subsystem that provisions integrated communication and compute resources on a per-subscriber basis, ensuring the coupled migration of both dimensions to maintain service continuity during mobility. SubEdge contributes the computing context--a per-subscriber data structure binding a Subscription Permanent Identifier (SUPI) to its inference container, edge node, and service entitlement--and a mobility-event-driven mechanism that simultaneously migrates the subscriber's compute instance and its traffic-routing policy when the serving cell changes. SubEdge operates as an Application Function over existing Network Exposure Function (NEF) APIs with zero 3GPP core modifications. Experimental evaluation on a real-world testbed shows that SubEdge's mobility-driven joint communication-and-compute migration reduces 95th-percentile latency from 22.9 ms to 12.2 ms with zero packet loss across six mobility events, sustains 99.92% frame delivery for an end-to-end 30 fps inference workload, and completes 1,560 migration operations across batches of up to 50 simultaneously migrating subscribers with 100% success.
