[A-7] Information Security

2:15 PM - 2:30 PM JST(5:15 AM - 5:30 AM UTC)Candidate for Academic Incentive Award

[A-7-05] A Fundamental Study on RFF Device Authentication Using Real-World Datasets

◎Kenta Ogawa¹, Ryoma Tanigawa¹, Yuhei Takahashi¹, Tomotaka Kimura¹, Jun Cheng¹ (1. Doshisha University)

Keywords:

radio-frequency fingerprint,authentication,physical-layer

In recent years, with the proliferation of IoT devices, physical-layer authentication using radio frequency fingerprints (RFFs)—hardware-specific characteristics that are difficult to tamper with—has garnered significant attention. Although authentication methods that do not require access lists and utilize RFF tags generated by the devices themselves have been proposed, previous studies were limited to simulations under restricted conditions, such as AWGN environments. Therefore, in this paper, we conduct a fundamental study of RFF device authentication using the large-scale real-signal dataset “WiSig” to verify its effectiveness in real-world environments where fading and noise coexist. We demonstrate the effectiveness of this method in real-world environments through authentication evaluations against impersonation attacks by unauthorized devices and an analysis of the distance distribution of features.