A Ray-Tracing Based Fingerprinting Method for Passive Localization in Urban NLOS Environment

Non-line-of-sight (NLOS) error mitigation for localization in urban environment has been a challenging problem because NLOS errors are site-specific and difficult to model. RF feature maps, which provide location specific RF signatures, can be a promising solution to overcome the NLOS problems. In this paper, we present an RF fingerprinting-based localization approach with the aid of ray-tracing simulation. Our method does not make any assumption on the NLOS error statistics. Instead, it leverages knowledge of RF propagation for the area of interest, accounting for all the buildings and terrains. Some key advantages of our method include 1) it only requires a simulated RF feature database and does not require intensive site surveys, and 2) it demonstrates superior localization performance in NLOS conditions compared with the existing techniques, which is validated through a real experiment conducted in an urban environment.