A Robust Zero-Calibration RF-based Localization System for Realistic Environments
Due to the noisy indoor radio propagation channel, Radio Frequency (RF)-based location determination systems usually require a tedious calibration phase to construct an RF fingerprint of the area of interest. This fingerprint varies with the used mobile device, changes of the transmit power of smart access points (APs), and dynamic changes in the environment; requiring re-calibration of the area of interest; which reduces the technology ease of use. In this paper, we present IncVoronoi: a novel system that can provide zero-calibration accurate RF-based indoor localization that works in realistic environments. The basic idea is that the relative relation between the received signal strength from two APs at a certain location reflects the relative distance from this location to the respective APs. Building on this, IncVoronoi incrementally reduces the user ambiguity region based on refining the Voronoi tessellation of the area of interest. IncVoronoi also includes a number of modules to efficiently run in realtime as well as to handle practical deployment issues including the noisy wireless environment, obstacles in the environment, heterogeneous devices hardware, and smart APs. We have deployed IncVoronoi on different Android phones using the iBeacons technology in a university campus. Evaluation of IncVoronoi with a side-by-side comparison with traditional fingerprinting techniques shows that it can achieve a consistent median accuracy of 2.8m under different scenarios with a low beacon density of one beacon every 44m2. Compared to fingerprinting techniques, whose accuracy degrades by at least 156%, this accuracy comes with no training overhead and is robust to the different user devices, different transmit powers, and over temporal changes in the environment. This highlights the promise of IncVoronoi as a next generation indoor localization system.