FedAPTA: Federated Multi-task Learning in Computing Power Networks with Adaptive Layer-wise Pruning and Task-aware Aggregation

2025-08-04

AI Summary

To address computational resource waste and model-structure mismatch arising from device and data heterogeneity in Computing Power Networks (CPNs) under federated multitask learning, this paper proposes a hierarchical adaptive pruning and task-aware heterogeneous aggregation framework. Methodologically: (1) we design a task-driven hierarchical pruning strategy that dynamically trims task-specific subnetworks per device according to its computational capacity; (2) we introduce a heterogeneous model recovery and aggregation mechanism based on task clustering, enabling efficient fusion of locally trained models with significantly divergent architectures. Extensive experiments on a real-world federated learning platform demonstrate that our approach consistently outperforms nine state-of-the-art baselines, achieving up to a 4.23% absolute accuracy gain. It substantially improves multitask convergence speed, computational resource utilization, and cross-device generalization capability.

Technology Category

• Machine Learning: Distributed Machine Learning & Federated Learning
• Constraint Satisfaction and Optimization: Distributed CSP/Optimization
• Multiagent Systems: Adversarial Agents

Application Category

• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Systems and Infrastructure for Web, Mobile and WoT: Federated Web and WoT systems, including distributed, federated and edge-based data processing
• Search and Retrieval-Augmented AI: Efficiency and scalability of Web search engines

Abstract

Federated Learning (FL) has shown considerable promise in Computing Power Networks (CPNs) for privacy protection, efficient data utilization, and dynamic collaboration. Although it offers practical benefits, applying FL in CPNs continues to encounter a major obstacle, i.e., multi-task deployment. However, existing work mainly focuses on mitigating FL's computation and communication overhead of a single task while overlooking the computing resource wastage issue of heterogeneous devices across multiple tasks in FL under CPNs. To tackle this, we design FedAPTA, a federated multi-task learning framework in CPNs. FedAPTA alleviates computing resource wastage through the developed layer-wise model pruning technique, which reduces local model size while considering both data and device heterogeneity. To aggregate structurally heterogeneous local models of different tasks, we introduce a heterogeneous model recovery strategy and a task-aware model aggregation method that enables the aggregation through infilling local model architecture with the shared global model and clustering local models according to their specific tasks. We deploy FedAPTA on a realistic FL platform and benchmark it against nine SOTA FL methods. The experimental outcomes demonstrate that the proposed FedAPTA considerably outperforms the state-of-the-art FL methods by up to 4.23%. Our code is available at https://github.com/Zhenzovo/FedCPN.

Problem

• Addresses multi-task deployment challenges in Federated Learning for CPNs
• Reduces computing resource wastage via adaptive layer-wise pruning
• Enables task-aware aggregation of heterogeneous local models

Innovation

• Adaptive layer-wise pruning for resource efficiency
• Task-aware aggregation for heterogeneous models
• Heterogeneous model recovery strategy integration