Open Set RF Fingerprinting Identification: A Joint Prediction and Siamese Comparison Framework

TL;DR

This work tackles open-set RF fingerprinting for device authentication in wireless IoT, where unknown rogue devices must be detected without prior examples. It introduces JRFFP-SC, a two-stage framework that first uses a VGG11-based RF fingerprint predictor (RFFP) to predict a probable identity and then employs a Siamese comparison (SIA) to verify the prediction against registered exemplars, mitigating inter-class interference. On a LoRa CSS dataset with 45 devices, JRFFP-SC achieves 98.47% legitimate accuracy and an AUC of 0.979 with an EER of 0.061 for rogue detection, outperforming a strong baseline and a purely SIA-based approach, and shows robustness to varying SNRs. The approach enables reliable open-set RF fingerprinting suitable for practical, low-overhead device authentication and rogue-device detection in wireless networks.

Abstract

Radio Frequency Fingerprinting Identification (RFFI) is a lightweight physical layer identity authentication technique. It identifies the radio-frequency device by analyzing the signal feature differences caused by the inevitable minor hardware impairments. However, existing RFFI methods based on closed-set recognition struggle to detect unknown unauthorized devices in open environments. Moreover, the feature interference among legitimate devices can further compromise identification accuracy. In this paper, we propose a joint radio frequency fingerprint prediction and siamese comparison (JRFFP-SC) framework for open set recognition. Specifically, we first employ a radio frequency fingerprint prediction network to predict the most probable category result. Then a detailed comparison among the test sample's features with registered samples is performed in a siamese network. The proposed JRFFP-SC framework eliminates inter-class interference and effectively addresses the challenges associated with open set identification. The simulation results show that our proposed JRFFP-SC framework can achieve excellent rogue device detection and generalization capability for classifying devices.

Figures (7)

• Figure 1: Illustration of legitimate device feature overlap.
• Figure 2: Proposed JRFFP-SC framework.
• Figure 3: The separate network structures of proposed JRFFP-SC framework.
• Figure 4: Training and inference illustrations of proposed JRFFP-SC framework.
• Figure 5: Classification results for 30 legal categories.

Theorems & Definitions (1)

• Remark 1