Adaptive Joint Compression and Synchronisation in Federated Split Learning for IoT Rainfall Prediction

2026-06-23

AI Summary

This work addresses the communication bottleneck in federated split learning caused by frequent activation uploads and gradient synchronization on IoT devices, specifically targeting rainfall prediction tasks. The authors propose a novel federated split learning framework that jointly optimizes activation compression and synchronization intervals. A server-side latency-aware adaptive scheduler and a client-side exponential moving average (EMA) smoothing mechanism are introduced to dynamically balance communication efficiency and model stability. Real-world deployment on Raspberry Pi demonstrates an 87% reduction in activation upload volume and a 54% decrease in synchronization traffic compared to the float32 baseline. Moreover, runtime jitter is reduced from ±688 seconds to ±10 seconds, while AUPRC variation remains below 0.011, achieving a significant trade-off among communication overhead, timing stability, and prediction accuracy.

Abstract

Federated split learning (FSL) enables collaborative training across bandwidth-constrained IoT devices, but repeated activation and gradient exchange creates a communication bot-tleneck. Prior work optimises either activation compression or synchronisation frequency in isolation. This paper presents an FSL framework for IoT rainfall prediction that jointly regulates activation compression and the synchronisation interval \r{ho} via a latency driven scheduler on a server with per client EMA smoothing. The system is evaluated on hourly ERA5 data from 11 weather stations through a 17 scenario simulation matrix and a four scenario Raspberry Pi deployment over a real wide-area link. The simulation matrix validates scheduler switching across low, high, and mixed latency profiles, while the Pi deployment validates the high latency endpoint selected by the same policy. AUPRC varies only slightly across configurations (0.6381-0.6484 in simulation; within 0.011 on Pi), indicating that aggressive quantisation and sparser aggregation do not materially degrade predictive quality in this setting. On Pi, the selected endpoint (int8 with rho=3) achieves an 87% reduction in activation upload payload and a 54% reduction in synchronisation traffic relative to the float32 baseline, while reducing runtime jitter from +/-688 s to +/-10 s.

Problem

• Federated Split Learning
• IoT
• Communication Bottleneck
• Activation Compression
• Synchronisation

Innovation

• federated split learning
• adaptive compression
• synchronisation scheduling
• latency-aware optimization
• IoT rainfall prediction