P3SL: Personalized Privacy-Preserving Split Learning on Heterogeneous Edge Devices

2025-07-23

AI Summary

Existing Split Learning (SL) frameworks struggle to support personalized model partitioning and local training across heterogeneous edge devices, owing to significant disparities in computational capability, communication bandwidth, environmental dynamics, and privacy requirements. Method: This paper proposes P3SL—a Personalized, Privacy-preserving Split Learning framework—that introduces a customizable two-level optimization mechanism. It enables each client to autonomously determine its optimal model split point and jointly optimize its local training strategy—without sharing raw data or sensitive intermediate features. Contribution/Results: P3SL supports diverse model architectures and adapts dynamically to changing environments. Evaluated on a 7-node heterogeneous testbed, it simultaneously achieves high model accuracy, significantly reduced privacy leakage risk, and lower energy consumption. To the best of our knowledge, P3SL is the first framework to enable adaptive collaborative learning under the triple heterogeneity of resource constraints, environmental volatility, and privacy requirements.

Technology Category

• Machine Learning: Learning on the Edge & Model Compression
• Search and Optimization: Learning to Search
• Natural Language Processing: Learning & Optimization for NLP

Application Category

• User Modeling, Personalization and Recommendation: User privacy protection in personalized systems
• Security and Privacy: Privacy-enhancing technologies
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Split Learning (SL) is an emerging privacy-preserving machine learning technique that enables resource constrained edge devices to participate in model training by partitioning a model into client-side and server-side sub-models. While SL reduces computational overhead on edge devices, it encounters significant challenges in heterogeneous environments where devices vary in computing resources, communication capabilities, environmental conditions, and privacy requirements. Although recent studies have explored heterogeneous SL frameworks that optimize split points for devices with varying resource constraints, they often neglect personalized privacy requirements and local model customization under varying environmental conditions. To address these limitations, we propose P3SL, a Personalized Privacy-Preserving Split Learning framework designed for heterogeneous, resource-constrained edge device systems. The key contributions of this work are twofold. First, we design a personalized sequential split learning pipeline that allows each client to achieve customized privacy protection and maintain personalized local models tailored to their computational resources, environmental conditions, and privacy needs. Second, we adopt a bi-level optimization technique that empowers clients to determine their own optimal personalized split points without sharing private sensitive information (i.e., computational resources, environmental conditions, privacy requirements) with the server. This approach balances energy consumption and privacy leakage risks while maintaining high model accuracy. We implement and evaluate P3SL on a testbed consisting of 7 devices including 4 Jetson Nano P3450 devices, 2 Raspberry Pis, and 1 laptop, using diverse model architectures and datasets under varying environmental conditions.

Problem

• Addresses split learning challenges in heterogeneous edge environments
• Customizes privacy and local models for diverse device conditions
• Optimizes split points without sharing sensitive client data

Innovation

• Personalized split learning for heterogeneous edge devices
• Bi-level optimization for privacy and resource management
• Customized privacy protection and local model maintenance