Device-free Movement Tracking using the UWB Channel Impulse Response with Machine Learning

Wireless communications systems are increasingly used for environmental sensing in addition to their main purpose of transmitting information. One way to use wireless communications systems for sensing is by using the channel impulse response (CIR) which captures the physical environment. Ultra-wideband (UWB) systems have a high-resolution CIR due to their large bandwidth, making them particularly attractive for sensing purposes, especially for device-free localization tasks. In this work, we use the temporary variation of the CIR on different delay bins over a time window as features in conjunction with machine learning techniques to detect the movement position and direction of people in an indoor environment.