LAUSR

As an important part of the Internet of Things, WiFi-based fingerprint indoor localization technology has been extensively studied due to the widespread deployment of WLAN in indoor environments. Nevertheless, there are two stubborn issues in indoor localization using the fingerprint method, one is the additional localization error caused by heterogeneous devices and the other is the huge time consumed by upgrading the fingerprint database. Here, we propose a dual-frequency difference and cubic spline interpolation (DFD-CSI) localization system in this article. First, according to the characteristics of commercial WiFi router generally supports 2.4- and 5-GHz dual bands, a fingerprint database is established using these two bands’ data and their difference data. The existence of difference solves the problem of device heterogeneity, while the existence of dual-frequency data makes the positioning result more robust. In addition, we also use cubic spline interpolation to upgrade the rough fingerprint database in order to reduce the time of site exploration on the premise of ensuring localization accuracy. Extensive experiments have been conducted in a campus hall. Results show that DFD-CSI is robust against altered devices and access points (APs) compared with the traditional heterogeneous devices algorithms and crowdsourcing update methods, achieving a mean error of 3.7 m and 27.5% localization error reduction with APs changed.