InfRA-FL: Information-Driven Robust Federated Learning via Saturation-Aware Reinforcement Learning
Machine Learning and Knowledge Extraction · 2026-08-14 · 8 (8) · 247
DOI: 10.3390/make8080247 ↗At a glance
Journal article in Machine Learning and Knowledge Extraction, published online on 2026-08-14. Written by Jiao Tian and 2 co-authors. 25 references. The full text is freely available (open access, CC BY 4.0). Funded by: the Natural Science Foundation of Xinjiang Uyghur Autonomous Region, Scientific Research Foundation of Xinjiang University for Young Scholars.
Abstract (original language)
Client selection is a critical mechanism for ensuring robust convergence in Federated Learning (FL) systems, yet it remains vulnerable to Non-IID data distributions and Byzantine attacks. Deep Reinforcement Learning (DRL) has shown promise for automated client selection, yet existing methods suffer from three structural deficiencies: observation ambiguity, where scalar states cannot distinguish malicious updates from benign heterogeneity; reward fragility, whereby attackers exploit unbounded feedback to hijack policy updates; and risk blindness, as risk-neutral agents overlook the inherent variance in client contributions. To address these deficiencies, we propose InfRA-FL, a robust adaptive framework. A mutual-information-based state construction extracts high-utility features, resolving observation ambiguity. A Saturation-Aware Robust Reward (SARR) mechanism applies soft-clipping to bound each client’s influence on the policy gradient, provably neutralizing reward poisoning. URA-PPO, an uncertainty-aware algorithm with a dual-head Critic, optimizes a risk-penalized objective that shifts the agent from risk-neutral to risk-averse decision-making. Experiments on MNIST and CIFAR-10 under 20% Byzantine adversaries show that InfRA-FL outperforms state-of-the-art baselines by 5–10% in accuracy while accelerating convergence, establishing that principled information-theoretic observation and robust reward design suffice to secure RL-driven federated learning against targeted poisoning.
Funding
- the Natural Science Foundation of Xinjiang Uyghur Autonomous Region
- Scientific Research Foundation of Xinjiang University for Young Scholars
Cite
APA
Tian, J., He, J., & Wang, L. (2026). InfRA-FL: Information-Driven Robust Federated Learning via Saturation-Aware Reinforcement Learning. Machine Learning and Knowledge Extraction, 8(8), 247. https://doi.org/10.3390/make8080247
Vancouver
Tian J, He J, Wang L. InfRA-FL: Information-Driven Robust Federated Learning via Saturation-Aware Reinforcement Learning. Machine Learning and Knowledge Extraction. 2026;8(8):247. doi:10.3390/make8080247
BibTeX
@article{Tian2026,
author = {Tian, Jiao and He, Jinlin and Wang, Liejun},
title = {InfRA-FL: Information-Driven Robust Federated Learning via Saturation-Aware Reinforcement Learning},
journal = {Machine Learning and Knowledge Extraction},
year = {2026},
volume = {8},
number = {8},
pages = {247},
doi = {10.3390/make8080247}
}
EUVIMED