pFedWN: A Personalized Federated Learning Framework for D2D Wireless Networks with Heterogeneous Data

TL;DR

This paper introduces pFedWN, a decentralized personalized federated learning framework that optimizes model performance over D2D wireless networks with heterogeneous data and channel conditions.

Contribution

The paper proposes a novel server-free PFL approach that incorporates wireless channel conditions and optimizes neighbor selection and weight assignment using EM, improving personalization and robustness.

Findings

pFedWN outperforms existing FL and PFL methods in diverse wireless scenarios.

It effectively handles non-IID and unbalanced datasets.

The method enhances learning robustness under dynamic wireless channels.

Abstract

Traditional Federated Learning (FL) approaches often struggle with data heterogeneity across clients, leading to suboptimal model performance for individual clients. To address this issue, Personalized Federated Learning (PFL) emerges as a solution to the challenges posed by non-independent and identically distributed (non-IID) and unbalanced data across clients. Furthermore, in most existing decentralized machine learning works, a perfect communication channel is considered for model parameter transmission between clients and servers. However, decentralized PFL over wireless links introduces new challenges, such as resource allocation and interference management. To overcome these challenges, we formulate a joint optimization problem that incorporates the underlying device-to-device (D2D) wireless channel conditions into a server-free PFL approach. The proposed method, dubbed pFedWN, optimizes the learning performance for each client while accounting for the variability in D2D wireless channels. To tackle the formulated problem, we divide it into two sub-problems: PFL neighbor selection and PFL weight assignment. The PFL neighbor selection is addressed through channel-aware neighbor selection within unlicensed spectrum bands such as ISM bands. Next, to assign PFL weights, we utilize the Expectation-Maximization (EM) method to evaluate the similarity between clients' data and obtain optimal weight distribution among the chosen PFL neighbors. Empirical results show that pFedWN provides efficient and personalized learning performance with non-IID and unbalanced datasets. Furthermore, it outperforms the existing FL and PFL methods in terms of learning efficacy and robustness, particularly under dynamic and unpredictable wireless channel conditions.