PARADIS: Physical 802.11 Device Identification with Radiometric Signatures

This paper describes a technique that identifies an ieee 802.11 frame’s source network interface card through passive radio-frequency analysis. Our approach, called paradis, leverages minute imperfections of transmitter hardware that are acquired at manufacture and are present even in otherwise identical nics. These imperfections are transmitterspecific and manifest themselves as artifacts of the emitted signals. We measure artifacts of individual wireless frames in the modulation domain, identify a suite of differentiating features, and apply efficient 802.11-specific machine-learning based classification techniques to achieve significantly higher degrees of accuracy than prior best known schemes. We experimentally demonstrate effectiveness of paradis in differentiating between more than a hundred 100 identical ieee 802.11 nics, with an accuracy in excess of 99.99%. Our results also show that the accuracy of paradis is resilient against ambient noise and changing characteristics of the wireless channel.