DOI: 10.1145/3832016 ISSN: 2474-9567

EcoTrack: Enabling Energy-Efficient Millimeter-Accurate Tracking on Commodity UWB Devices

Xinran Li, Junqi Ma, Fusang Zhang, Beihong Jin, Hua Zhong, Jie Xiong

The recent integration of ultra-wideband (UWB) radios into mobile devices has opened the door to a new wave of Internet of Things (IoT) applications. While prior work has primarily focused on improving tracking accuracy, energy constraints in battery-powered mobile devices remain a persistent challenge. We present EcoTrack, a novel uplink phase-based UWB tracking system that significantly reduces energy consumption while still achieving millimeter-level accuracy. EcoTrack optimizes energy consumption by fundamentally redesigning conventional UWB tracking methods through two key strategies: (i) reducing the number of message transmissions and (ii) reducing location update rates. Specifically, departing from conventional methods that rely on multiple message exchanges, we propose an uplink scheme that requires the mobile device to broadcast only a single message for each location update. For accurate phase estimation under unsynchronized multi-anchor scenarios, we propose a two-stage differential operation to cancel the time-varying phase errors in raw phase estimates. While lowering the update rate effectively reduces energy consumption, it also introduces phase ambiguity for target tracking. To tackle this challenge, we fuse wrapped UWB phase with low-power IMU readings widely available on mobile devices to achieve robust ambiguity resolution. Extensive experiments demonstrate that EcoTrack reduces energy consumption by up to 84% compared to existing methods while maintaining a low median 3D tracking error of 0.78 cm across diverse indoor environments.