Federated Learning for LLMs over Mobile Networks: Issues and Solutions in the RAN Transport

2026-10-01

AI Summary

This study addresses the challenge of efficiently allocating transmission resources during asynchronous federated fine-tuning of large models in mobile networks by proposing a learning-aware traffic shaping architecture. Leveraging a radio access network (RAN)-aware learning framework, the proposed method employs in-network aggregation at base stations to transform asynchronous updates into deterministic traffic, thereby harmonizing wireless access flexibility with the predictable scheduling of optical circuit switching. This work pioneers the integration of AI training workflows with programmable optical transport networks, successfully reshaping unpredictable federated learning traffic into patterns amenable to on-demand optical connection provisioning. Consequently, the proposed approach significantly enhances overall network resource utilization.

Abstract

Federated LLM fine-tuning enables large models to be adapted using private and geographically distributed data at the network edge, creating recurring and deadline-sensitive communication workloads across access and transport networks. This challenge is particularly relevant in mobile RANs, where wireless variability, mobility, and device heterogeneity cause model updates to arrive asynchronously. Although these updates belong to the same learning round and share a common destination and deadline, conventional transport networks treat them as independent device-originated flows, hiding their underlying structure and limiting the ability to efficiently provision transport resources. This mismatch is particularly problematic for optical circuit switching and all-photonics transport, which benefit from predictable and schedulable traffic demands. We argue that future RANs should act as learning-aware traffic shapers by exposing the communication structure of distributed model adaptation to the transport layer. Through in-network aggregation at the gNB, asynchronous UE updates can be transformed into fewer aggregate transfers with bounded size and delivery requirements. Once shaped in this way, federated LLM traffic becomes a suitable candidate for selectively provisioned optical connectivity, where high-capacity paths can be established during aggregate-transfer windows and released between learning rounds. The resulting architecture combines the flexibility of packet-based mobile access with dynamically provisioned optical capacity, illustrating a broader approach for coordinating distributed AI workloads across programmable access and transport networks.

Problem

• Federated Learning
• Large Language Models
• Mobile RAN
• Optical Transport
• Asynchronous Updates

Innovation

• Federated Learning
• Large Language Models
• In-network Aggregation
• Optical Circuit Switching
• Radio Access Network