FedSkipTwin: Digital-Twin-Guided Client Skipping for Communication-Efficient Federated Learning

2025-07-17

AI Summary

To address excessive communication overhead in bandwidth-constrained edge devices (e.g., mobile terminals and IoT nodes) under federated learning, this paper proposes a client-adaptive skipping mechanism leveraging a lightweight server-side digital twin. Innovatively, it deploys an LSTM-driven digital twin model on the server to jointly predict client update magnitudes and epistemic uncertainty by modeling historical gradient norm sequences. Based on these predictions, the server dynamically decides whether to skip the current local model upload. The mechanism incorporates a threshold-adaptation strategy, achieving 12–15.5% communication reduction under non-IID data while improving model accuracy by up to 0.5 percentage points—outperforming FedAvg significantly. This approach effectively balances communication efficiency and model performance.

Technology Category

• Machine Learning: Distributed Machine Learning & Federated Learning
• Search and Optimization: Learning to Search
• Natural Language Processing: Learning & Optimization for NLP

Application Category

• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Systems and Infrastructure for Web, Mobile and WoT: Federated Web and WoT systems, including distributed, federated and edge-based data processing

Abstract

Communication overhead remains a primary bottleneck in federated learning (FL), particularly for applications involving mobile and IoT devices with constrained bandwidth. This work introduces FedSkipTwin, a novel client-skipping algorithm driven by lightweight, server-side digital twins. Each twin, implemented as a simple LSTM, observes a client's historical sequence of gradient norms to forecast both the magnitude and the epistemic uncertainty of its next update. The server leverages these predictions, requesting communication only when either value exceeds a predefined threshold; otherwise, it instructs the client to skip the round, thereby saving bandwidth. Experiments are conducted on the UCI-HAR and MNIST datasets with 10 clients under a non-IID data distribution. The results demonstrate that FedSkipTwin reduces total communication by 12-15.5% across 20 rounds while simultaneously improving final model accuracy by up to 0.5 percentage points compared to the standard FedAvg algorithm. These findings establish that prediction-guided skipping is a practical and effective strategy for resource-aware FL in bandwidth-constrained edge environments.

Problem

• Reduces communication overhead in federated learning
• Improves model accuracy with client-skipping strategy
• Optimizes bandwidth usage for edge devices

Innovation

• Digital-twin-guided client skipping algorithm
• LSTM-based server-side gradient prediction
• Threshold-based communication skipping strategy