Review of radio frequency fingerprinting identification

The traditional wireless network security based on cryptographic mechanism and security protocol has potential safety hazard. The new physical layer-based RF fingerprinting method uses the fine-feature of transmitter RF signals to distinguish different individuals. It has the advantages of being difficult to clone and forge, and has broad application prospects. This paper first discusses the four basic characteristics of ideal RF fingerprinting, namely uniqueness, time invariance, independence and robustness, and analyzes the research status in four basic characteristics. Then according to the three parts of signal preprocessing, feature extraction and classification, the status of RF fingerprint identification is summarized. The RF-distinct native attribute(RF-DNA), modulation domain and RF fingerprinting method based on deep learning are analyzed. Finally, the various signal types/modulation methods and corresponding application scenarios involved in RF fingerprint identification research are summarized, and the broad application prospects of RF fingerprint recognition are presented. The research trend of RF fingerprint recognition is discussed.