MagneticSpy: Exploiting Magnetometer in Mobile Devices for Website and Application Fingerprinting

Abstract

Recent studies have shown that aggregate CPU usage and power consumption traces on smartphones can leak information about applications running on the system or websites visited. In response, access to such data has been blocked for mobile applications starting from Android 7. In this work, we explore a new source of sidechannel leakage for this class of a acks. Our method is based on the fact that electromagnetic activity caused by mobile processors leads to noticeable disturbances in magnetic sensor measurements on mobile devices, with the amplitude being proportional to the CPU workload. erefore, recorded sensor data can be analyzed to reveal information about ongoing activities. e a ack works on a number of devices: We evaluated 59 models of modern smartphones and tablets and observed the reaction of the magnetometer to CPU activity on 39 of them. On selected devices we were able to successfully identify which application has been opened (with up to 90% accuracy) or which web page has been loaded (up to 91% accuracy). We believe that the presented side channel poses a significant risk to end users' privacy, as the sensor data can be recorded from native apps and even from web pages without user permissions. Finally, we discuss possible countermeasures to prevent the presented information leakage.

S3

“…Block et al leveraged magnetometers' ability to detect locations of users' devices within commercial GPS accuracy [7]. Researchers also demonstrated that electromagnetic field measured by smartphone magnetometers could be exploited for webpage [22], and device [29] fingerprinting. Compared with existing side channels, we introduce a new side-channel attack to identify users' handwriting by analyzing the magnetic impact of the embedded magnets in modern stylus pencils sensed by device magnetometers.…”

“…Prior efforts have unveiled security and privacy concerns that result from applications having access to motion sensors' data. Such works propose attacks targeted at inferring users' privacy-sensitive data such as touch actions and keystrokes [21,24,33], passwords and PIN codes [24,28], and application and webpage fingerprinting [22,27]. Unfortunately, side-channel attacks continue to be a major source of concern to users and developers due to frequent software updates and new devices introduced in the market.…”