Engineers prove for the first time the ability to track devices using Bluetooth signals

A group of engineers from the University of California San Diego for the first time provedthat the Bluetooth signals constantly emitted by our mobile phones and accessories have a unique footprint that can be used to track a person's movement.

Experts point out that all wireless devices have small hardware manufacturing defects that are unique to each device. These "fingerprints", being random by-products of the manufacturing process, result in unique distortions that can be used to trace a particular device.

Some previous studies have shown that a WiFi signal generates a similar trace. To identify it, a long sequence called the preamble is used. But the same method is not suitable for identifying a Bluetooth signal, because its preamble is too short.

For this reason, researchers have developed a new method that does not rely on the preamble, but considers the entire Bluetooth signal. They developed an algorithm that evaluates two different values ​​in its signals. These values ​​vary with hardware defects, giving researchers a unique device footprint.

In real tests, this method showed that 40% of the 162 mobile devices seen in public places, such as cafes, were uniquely identifiable. The experiment was then expanded and 647 mobile devices were monitored for two days. Unique "fingerprints" were identified in 47%.

Finally, engineers have demonstrated the ability to track the movements of a user who has become a volunteer. The unique Bluetooth footprint made it possible to record his movements as he entered and left the house.

The researchers note that their method has a number of problems that attackers who want to put the method into practice will face. For example, changes in ambient temperature can change the unique Bluetooth footprint. Also, some devices can send signals of different strength, which greatly affects the distance at which these devices can be identified.

In addition, this method requires a highly skilled attacker, so today it is unlikely to pose a widespread threat to mobile device users.

Currently, a team of engineers is working on learning how to hide the Bluetooth trace using digital signal processing in the device's firmware. This could eliminate surveillance if implemented everywhere.