A Novel Indoor Human Occlusion Identification Method With BLE Signals

The continuous development of the Internet of Things (IoT) and the Industrial Internet has given rise to many new application scenarios and demands, which put forward new demands and challenges for wireless short-range communication technologies. Among them, Bluetooth Low Energy (BLE) stands out in indoor single-station ranging by virtue of its low cost and low power. However, in the indoor environment, BLE signal propagation is easily occluded by the human body to form none-line-of-sight (NLOS) propagation which affects the range accuracy. In this work, we explore the difference in the amplitude and phase information of the channel response in the presence or absence of human occlusion, using frequency hopping and multi-packet transmission, and then we propose an identification method based on the coefficient of variation (COV), which can effectively identify the presence of human occlusion. Experimental results in indoor scenes show that our new features have higher identification accuracy than previous features.