Fingerprint and assistant nodes based Wi-Fi localization in complex indoor environment

With the extensive development of Wi-Fi, indoor location services based on received signal strength (RSS) fingerprints have attracted increasing attention from researchers. In complex indoor environments, multipath and non-line-of-sight (NLOS) conditions would lead to large errors in measured values, thereby reducing indoor positioning accuracy. In this paper, we propose a Wi-Fi indoor localization method based on collaboration of fingerprint and assistant nodes. First, appropriate assistant nodes based on the similarity of RSS sequences are elaborately selected around the unknown node and distances between them are used as auxiliary information to improve the positioning accuracy. Furthermore, in the complex indoor circumstances that result in NLOS error, an adaptive Kalman filter with colored noise is used to mitigate the time-of-flight ranging error. Experiments demonstrate that in complex indoor environments, our system can outperform its counterparts with robust performance and low localization estimation error.