Improvement of Radio Frequency Fingerprint Portability for Wi-Fi Adaptor Identification

Abstract

As quantum computing technology approaches maturity, the security of wireless communication using data encryption algorithms is facing great challenges. On the other hand, physical layer security technologies using the unique physical characteristic properties of one transmitter like radio frequency/ radiometric fingerprint, localization information, or channel state information are more resistant to attacks using quantum computing technology. Consequently, the research of physical layer security has attracted significant attentions recently. However, the RF fingerprint portability is an inevitable problem in practical applications. In this work, a joint RF feature calibration and weighted feature classification scheme is proposed to overcome the RF fingerprint portability problem caused by channel fading and receiver differentiation. Experimental results show that the average successful identification rate increases significantly using the proposed scheme.