Improving Federated Aggregation with Deep Unfolding Networks

TL;DR

This paper proposes a deep unfolding network-based method to improve federated learning aggregation by learning adaptive, unbiased weights that mitigate client heterogeneity effects, enhancing accuracy and interpretability.

Contribution

It introduces a novel DUN-based technique for adaptive, unbiased weighting in federated aggregation, addressing client heterogeneity more effectively than existing methods.

Findings

Achieves improved accuracy in heterogeneous FL environments.

Demonstrates effective and interpretable adaptive weighting.

Reduces computational power needed for aggregation.

Abstract

The performance of Federated learning (FL) is negatively affected by device differences and statistical characteristics between participating clients. To address this issue, we introduce a deep unfolding network (DUN)-based technique that learns adaptive weights that unbiasedly ameliorate the adverse impacts of heterogeneity. The proposed method demonstrates impressive accuracy and quality-aware aggregation. Furthermore, it evaluated the best-weighted normalization approach to define less computational power on the aggregation method. The numerical experiments in this study demonstrate the effectiveness of this approach and provide insights into the interpretability of the unbiased weights learned. By incorporating unbiased weights into the model, the proposed approach effectively addresses quality-aware aggregation under the heterogeneity of the participating clients and the FL environment. Codes and details are \href{https://github.com/shanikairoshi/Improved_DUN_basedFL_Aggregation}{here}.