Open-Set RF Fingerprinting via Improved Prototype Learning

TL;DR

This paper introduces an improved prototype learning approach for open-set RF fingerprinting, enabling recognition of unseen devices by enhancing feature robustness with regularization and label smoothing.

Contribution

It proposes two novel improvements—consistency-based regularization and online label smoothing—to adapt prototype learning for open-set RF fingerprinting.

Findings

Significantly improved open-set recognition performance

Effective handling of unseen device signals

Robust feature space learning

Abstract

Deep learning has been widely used in radio frequency (RF) fingerprinting. Despite its excellent performance, most existing methods only consider a closed-set assumption, which cannot effectively tackle signals emitted from those unknown devices that have never been seen during training. In this letter, we exploit prototype learning for open-set RF fingerprinting and propose two improvements, including consistency-based regularization and online label smoothing, which aim to learn a more robust feature space. Experimental results on a real-world RF dataset demonstrate that our proposed measures can significantly improve prototype learning to achieve promising open-set recognition performance for RF fingerprinting.