Target Tracking using Signal Strength Differences for Long-Range IoT Networks

Radio based positioning or tracking solutions typically require wideband signals or phase coherent antennas. In this paper, we present a target tracking method based on received non-coherent signal strength differences (RSSDs) between antennas for outdoor Internet-of-things (IoT) scenarios. We introduce an RSSD model based on classical path-loss models. With known antenna patterns and antenna array geometries, the RSSD model enables direct mapping between RSSD and angle of arrival, without involving parameters like transmit power, path-loss coefficient, etc. The RSSD model is then exploited in a recursive Bayesian filtering method for target tracking where a particle filter-based implementation is used. The performance is evaluated using outdoor measurements in a low-power wide area network (LoRaWAN) based IoT system. Besides, we also investigate the potential of the RSSD model for AoA estimation. The experimental results show the capability of the proposed framework for real-time target/AoA tracking; reasonable accuracy is achieved even when using non-averaged RSS measurements and under non line-of-sight (NLoS) conditions. Furthermore, the non-coherent approach has low computational complexity, scales well, and is flexible to allow for different antenna array configurations.