Encoder-Sharing Hierarchical Federated Multi-Task Learning for VANETs

2026-09-28

AI Summary

Existing federated learning approaches for the Internet of Vehicles typically assume that vehicles train a single model, limiting their adaptability to heterogeneous perception tasks. To address this limitation, this work proposes a multi-task collaborative framework based on clustered hierarchical federated learning. The framework aggregates cross-task generalizable features through a globally shared encoder while retaining local decoders tailored to specific tasks, thereby enabling transferable representation learning across heterogeneous tasks under privacy-preserving constraints. Experimental results demonstrate that the proposed framework improves model accuracy by up to 24.0% and reduces communication rounds by a maximum of 69 (28.8%), significantly enhancing both the efficiency and performance of collaborative learning.

Abstract

Most federated learning frameworks for vehicular ad hoc networks assume that all vehicles collaboratively train a single model for a common task. This assumption limits their applicability to practical vehicular environments, where vehicles may perform heterogeneous but related perception tasks with different output spaces. This paper proposes encoder-sharing hierarchical multi-task federated learning (EN-HMTFL), which integrates cluster-based hierarchical federated learning with a globally shared encoder and vehicle-local decoders. EN-HMTFL enables vehicles performing different tasks to collaboratively learn a transferable feature representation while preserving their task-specific models locally. Only the encoder is exchanged and aggregated through the hierarchy, whereas raw data and local decoder parameters remain at the vehicles. The proposed framework is evaluated on the MNIST and GTSRB datasets in different vehicular scenarios. Across the evaluated scenarios, EN-HMTFL improves accuracy by up to 24.0% relative to the compared representation-sharing benchmark. In scenarios where EN-HMTFL converges earlier, the reduction reaches up to 69 communication rounds (28.8%).

Problem

• Federated Learning
• Vehicular Ad Hoc Networks
• Multi-Task Learning
• Heterogeneous Tasks

Innovation

• Hierarchical Federated Learning
• Multi-Task Learning
• Encoder-Sharing
• Vehicular Ad Hoc Networks
• Representation Transfer