One GPU to Snoop Them All: a Full-Band Bluetooth Low Energy Sniffer

Sniffing Bluetooth data sessions is considered a difficult task, because of the frequency-hopping channel access scheme this technology implements. In this paper we present a novel open-source sniffer that can monitor Bluetooth Low Energy (BLE) traffic on all channels in real time. The sniffer builds on a Software-Defined Radio (SDR) framework to capture the entire BLE spectrum and exploits Graphics Processing Unit (GPU) capabilities to channelize and process BLE traffic in real time. We show that our sniffer can easily and reliably detect active BLE connections, and infer their properties, including Access Address, CRC values and hopping sequences. From a general standpoint, we show that tracking many BLE data sessions at the same time becomes feasible even with relatively inexpensive equipment, as we are able to discover up to 24 simultaneous sessions within 80 ms on average.