Smartphones are increasingly augmented with sensors for a variety of purposes. In this paper, we show how magnetic field emissions can be used to fingerprint smartphones. Previous work on identiication rely on speciic characteristics that vary with the settings and components available on a device. This limits the number of devices on which one approach is effective. By contrast, all electronic devices emit a magnetic ield which is accessible either through the API or measured through an external device. We conducted an in-the-wild study over four months and collected mobile sensor data from 175 devices. In our experiments we observed that the electromagnetic field measured by the magnetometer identifies devices with an accuracy of 98.9%. Furthermore, we show that even if the sensor was removed from the device or access to it was discontinued, identiication would still be possible from a secondary device in close proximity to the target. Our findings suggest that the magnetic field emitted by smartphones is unique and fingerprinting devices based on this feature can be performed without the knowledge or cooperation of users.
Perez, B.a.M. (2019). Fatal attraction: Identifying mobile devices through electromagnetic emissions. ACM [10.1145/3317549.3319726].
Fatal attraction: Identifying mobile devices through electromagnetic emissions
Musolesi, M;
2019
Abstract
Smartphones are increasingly augmented with sensors for a variety of purposes. In this paper, we show how magnetic field emissions can be used to fingerprint smartphones. Previous work on identiication rely on speciic characteristics that vary with the settings and components available on a device. This limits the number of devices on which one approach is effective. By contrast, all electronic devices emit a magnetic ield which is accessible either through the API or measured through an external device. We conducted an in-the-wild study over four months and collected mobile sensor data from 175 devices. In our experiments we observed that the electromagnetic field measured by the magnetometer identifies devices with an accuracy of 98.9%. Furthermore, we show that even if the sensor was removed from the device or access to it was discontinued, identiication would still be possible from a secondary device in close proximity to the target. Our findings suggest that the magnetic field emitted by smartphones is unique and fingerprinting devices based on this feature can be performed without the knowledge or cooperation of users.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.