DeepRadioID Article Swipe

PDF

Related Concepts

Computer science Fingerprint (computing) Wireless Transmitter Fingerprint recognition Protocol (science) Feature extraction Authentication (law) Wireless Application Protocol Feature (linguistics) Wireless network Radio frequency Computer network Embedded system Artificial intelligence Telecommunications Computer security Channel (broadcasting) Linguistics Medicine Philosophy Alternative medicine Pathology

Francesco Restuccia , Salvatore D’Oro , Amani Al-Shawabka , Mauro Belgiovine , Luca Angioloni , Stratis Ioannidis , Kaushik Chowdhury , Tommaso Melodia ·

YOU? · · 2019 · Open Access · · DOI: https://doi.org/10.1145/3323679.3326503 · OA: W2937290901

Radio fingerprinting provides a reliable and energy-efficient IoT authentication strategy by leveraging the unique hardware-level imperfections imposed on the received wireless signal by the transmitter's radio circuitry. Most of existing approaches utilize hand-tailored protocol-specific feature extraction techniques, which can identify devices operating under a pre-defined wireless protocol only. Conversely, by mapping inputs onto a very large feature space, deep learning algorithms can be trained to fingerprint large populations of devices operating under any wireless standard.