InSecTT BLE Channel Sniff Dataset

This dataset provides Bluetooth Low Energy (BLE) measurements where a single BLE channel is observed with a Software Defined Radio (SDR) in close proximity to several BLE devices. Since BLE uses channel hopping, not every message is detected if only a single channel is observed. In general, we do not know the access address, the connection interval, or the hopping pattern of the connections [1]. This measurement set provides raw filtered and GFSK demodulated I/Q samples of the received BLE messages, as well as the already extracted measurements for the observed BLE connections including the header and measured timestamp. For each measurement set, we provide metadata with information about the BLE devices that are part of the setup. In BLE there are currently two different channel hop algorithms used [1]. The first one is the Channel Selection Algorithm 1 (CSA1), which was released with the first BLE specification, and the second is the Channel Selection Algorithm 2 (CSA2) which was implemented with the BLE version 5.0. This dataset provides measurements with BLE devices using both algorithms placed near the sniffer. The goal is to show that even though channel hopping is used, it is possible to reconstruct certain communication parameters and predict future channel access of the BLE connections by only passively listening to the channel. This dataset is a work from Silicon Austria Labs GmbH (SAL) and the Institute for Communications Engineering and RF-Systems (NTHFS) of the Johannes Kepler University (JKU) in Linz for the InSecTT project.