Learning to Sync at the Edge: Client-Side Adaptive Synchronization for Federated Learning
The scalability and performance of edge federated learning are fundamentally constrained by synchronization bottlenecks.
ProofPaper ↗
Key points
- We present a client-side reinforcement learning synchronization framework in which each client decides during training whether to upload, download, synchronize bidirectionally, or defer communication.
- The framework places a lightweight scheduler alongside the local model and formulates synchronization as a resource-aware control problem driven by local learning and system state.
- We study this framework on two representative edge workloads, spatiotemporal traffic forecasting and image classification.
- Across both settings, adaptive client-side synchronization improves the accuracy-cost trade-off, reducing system cost in traffic forecasting while maintaining competitive cost and improving final accuracy in image classification.
Sources (1)
- [1]Learning to Sync at the Edge: Client-Side Adaptive Synchronization for Federated LearningarXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 9, 07:41 PM
The scalability and performance of edge federated learning are fundamentally constrained by synchronization bottlenecks.
We present a client-side reinforcement learning synchronization framework in which each client decides during training whether to upload, download, synchronize bidirectionally, or defer communication.
Extractive summary: sentences quoted from the sources.
Before this
- Oct 9, 2026Visualizing the World Model in an Agent's Value Function
- Oct 9, 2026Dream to Remember: Continual Steering of Generalist Robot Policies with World Models
- Oct 9, 2026The Thinking Leakage in Reinforcement Learning for LLMs
- Oct 9, 2026Rethinking Relative Policy Differences Transfer of On-Policy Distillation
- Oct 9, 2026SimVTA: Simulation-Grounded Video-Trajectory Alignment for Driving World Action Models
- Oct 9, 2026AutoEnvScaling: Automating the Data Flywheel with Terminal Agents