WiFi and BLE Fingerprinting for Smartphone Proximity Detection

Abstract

Wireless devices such as smartphones, wireless bracelets or smartwatches are now often used by a single person. They can therefore be used to model human contact, proximity to each other, or social networking. In light of events such as the recent pandemic interactions or proximity between people have become increasingly important. In this context, this paper explores the limitations of a machine learning-based approach that detects the proximity of two devices (up to about two metres) based on WiFi and BLE (Bluetooth Low Energy) fingerprints of their radio environments. Specifically, we compare the use of a rudimentary set of two features and an extended, more complex set of features, exploring the use of separate classifiers that treat WiFi and BLE features separately. In addition, we investigate whether using only one of the two communication technologies for detection could lead to better results and the importance of radio propagation expertise in feature extraction. We found that using a more complex set of features that can be subjected to further feature selection procedures can provide a performance benefit of about 4.6 percentage points. In terms of the communication technologies used, our results also show that using BLE alone always leads to significantly worse results than using WiFi alone or WiFi and BLE together. The evaluation was carried out in three different radio environments, namely indoors in companies, outdoors on the street and in the staircase of residential building with many apartments.