Human Presence Detection Using Bluetooth Channel Sounding

Accidents involving children being left behind in vehicles have become a serious social issue both domestically and internationally. To prevent such incidents, various methods using devices such as weight sensors, cameras, and millimeter-wave radars have been proposed, some of which have already been put into practical use. However, these approaches face challenges such as high implementation costs and limited space for additional wiring inside vehicles. To address these issues, we focus on Bluetooth devices, which are widely installed in many vehicles, and investigate human detection using Bluetooth Channel Sounding. Channel Sounding is a technique primarily used for ranging, capable of obtaining high-precision phase information
across multiple channels. When an object that affects radio waves, such as a human, is present near a Bluetooth device, the phase information obtained through Channel Sounding is expected to exhibit variations. In this study, we leverage changes in phase information and apply machine learning for human detection.