STrack: Velocity Estimation Using Single Antenna WiFi Devices

• Published in: 2023 Asia Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA ASC) pp. 927 - 933
• Main Authors: Xu, Jian, Zhang, Dongheng, Li, Jiamu, Sun, Qibin, Chen, Yan
• Format: Conference Proceeding
• Language: English
• Published: IEEE 31.10.2023
• Subjects: Estimation; Frequency estimation; Information processing; Internet of Things; Sensors; Symbols; Target tracking
• ISSN: 2640-0103
• DOI: 10.1109/APSIPAASC58517.2023.10317401

As the Internet of Things (IoT) continues to evolve, WiFi devices are becoming increasingly ubiquitous in our everyday lives. This proliferation of devices has paved the way for innovative technologies such as device-free passive indoor tracking using WiFi devices. For device-free passive WiFi sensing, the Doppler Frequency Shift (DFS) extracted from fine-grained Channel State Information (CSI) phase has been widely adopted for estimating target motion velocity. Nevertheless, due to the presence of random phase offsets in CSI and the fact that many IoT devices have only one WiFi antenna, utilizing CSI phase for calculating DFS is usually unfeasible. In this paper, we present a novel Single Antenna WiFi Tracking (STrack) system that estimates DFS using CSI amplitude, eliminating the need for DFS symbols to directly participate in velocity calculation. Extensive experimental results indicate that STrack outperforms state-of-the-art indoor tracking systems, with a median velocity error of 1.49cm/s, and a median direction error of 12.41°, a median positioning error of 48.529cm.