Fairness-Aware Client Selection for Federated Learning

TL;DR

This paper introduces FairFedCS, a novel client selection method for federated learning that balances fairness and performance by dynamically adjusting client participation probabilities based on reputation and contribution, leading to improved fairness and accuracy.

Contribution

The paper proposes a Lyapunov optimization-based client selection algorithm that enhances fairness without sacrificing model performance in federated learning.

Findings

Achieves 19.6% higher fairness compared to state-of-the-art methods.

Attains 0.73% higher test accuracy on real-world datasets.

Provides opportunities for clients to redeem reputation after poor performance.

Abstract

Federated learning (FL) has enabled multiple data owners (a.k.a. FL clients) to train machine learning models collaboratively without revealing private data. Since the FL server can only engage a limited number of clients in each training round, FL client selection has become an important research problem. Existing approaches generally focus on either enhancing FL model performance or enhancing the fair treatment of FL clients. The problem of balancing performance and fairness considerations when selecting FL clients remains open. To address this problem, we propose the Fairness-aware Federated Client Selection (FairFedCS) approach. Based on Lyapunov optimization, it dynamically adjusts FL clients' selection probabilities by jointly considering their reputations, times of participation in FL tasks and contributions to the resulting model performance. By not using threshold-based reputation filtering, it provides FL clients with opportunities to redeem their reputations after a perceived poor performance, thereby further enhancing fair client treatment. Extensive experiments based on real-world multimedia datasets show that FairFedCS achieves 19.6% higher fairness and 0.73% higher test accuracy on average than the best-performing state-of-the-art approach.