Dataset Collection of Multi-Communication Technologies Monitored in Different Mobility Contexts
Résumé
The ubiquitous nature of mobile devices equipped
with radio communication technologies made the collection of
data a commonplace especially for studying human mobility. The
collection of such datasets forms the intermediate results in many
scientific research projects. Therefore, with the lack of datasets,
collecting and publishing data should be seriously addressed
since several scientific research is based on the gathering and
analysis of measurement data. In this paper, we introduce the
PILOT dataset, a Privacy-preserving data collectIon of wireLess
cOmmunication Technologies. The dataset is a collection of four
jointly collected information in different mobility contexts. It
includes three wireless communication technologies: WiFi proberesponses, BLE (Bluetooth Low Energy) beacons, and LoRa (Long
Range Radio) packets, plus additional information: Acceleration,
Roll, and Pitch, all collected at the same time. We provide the
keys to reproduce such data collection and share the datasets
already collected. The dataset is collected for approximately 90
hours, with a size of 200 MB using FiPy devices from Pycom
and it is uploaded to GitHub. The dataset’s utility is validated
through the application of a classification machine learning model
that determines the real-life situation of devices through the
communication links monitored in different scenarios with an
accuracy of 94%. Thus, we believe that such dataset is important
for human mobility studies and applications of integrated sensing
systems since it offers a new form of a classified collected data
that does not exist in the already published datasets.