Adversarial Client Detection via Non-parametric Subspace Monitoring in the Internet of Federated Things

TL;DR

This paper introduces FedRR, a non-parametric method that detects adversarial clients in federated learning networks by monitoring low-rank features of parameter updates, enhancing security in the Internet of Federated Things.

Contribution

The paper presents a novel non-parametric approach, FedRR, for detecting adversarial clients in federated learning by analyzing low-rank features of updates, with proven effectiveness.

Findings

Accurately detects adversarial clients

Controls false alarm rate effectively

Validated on MNIST digit recognition dataset

Abstract

The Internet of Federated Things (IoFT) represents a network of interconnected systems with federated learning as the backbone, facilitating collaborative knowledge acquisition while ensuring data privacy for individual systems. The wide adoption of IoFT, however, is hindered by security concerns, particularly the susceptibility of federated learning networks to adversarial attacks. In this paper, we propose an effective non-parametric approach FedRR, which leverages the low-rank features of the transmitted parameter updates generated by federated learning to address the adversarial attack problem. Besides, our proposed method is capable of accurately detecting adversarial clients and controlling the false alarm rate under the scenario with no attack occurring. Experiments based on digit recognition using the MNIST datasets validated the advantages of our approach.