FedTAD: Topology-aware Data-free Knowledge Distillation for Subgraph Federated Learning

TL;DR

FedTAD introduces a topology-aware, data-free knowledge distillation method to improve subgraph federated learning by addressing heterogeneity in node and topology variations, leading to more reliable global GNNs.

Contribution

This paper proposes FedTAD, a novel topology-aware, data-free knowledge distillation approach that effectively mitigates subgraph heterogeneity in federated GNN training.

Findings

FedTAD outperforms state-of-the-art baselines on six datasets.

Decoupling node and topology variations reveals their impact on knowledge reliability.

Topology-aware distillation improves global GNN performance in heterogeneous subgraph settings.

Abstract

Subgraph federated learning (subgraph-FL) is a new distributed paradigm that facilitates the collaborative training of graph neural networks (GNNs) by multi-client subgraphs. Unfortunately, a significant challenge of subgraph-FL arises from subgraph heterogeneity, which stems from node and topology variation, causing the impaired performance of the global GNN. Despite various studies, they have not yet thoroughly investigated the impact mechanism of subgraph heterogeneity. To this end, we decouple node and topology variation, revealing that they correspond to differences in label distribution and structure homophily. Remarkably, these variations lead to significant differences in the class-wise knowledge reliability of multiple local GNNs, misguiding the model aggregation with varying degrees. Building on this insight, we propose topology-aware data-free knowledge distillation technology (FedTAD), enhancing reliable knowledge transfer from the local model to the global model. Extensive experiments on six public datasets consistently demonstrate the superiority of FedTAD over state-of-the-art baselines.