.png){kind=icon}
.png){kind=icon}
Return
Indoor ranging localization algorithm using LOSAPs for smartphones
.png){kind=icon}
PRE
AI
DOI：10.1016/j.comnet.2025.111920.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
This paper presents an indoor ranging localization algorithm for smartphones using Line-of-Sight Access Points (LOSAPs), enabling WiFi Round-Trip Time (RTT)-based indoor positioning. The proposed algorithm consists of three key stages: LOSAP search, ranging calibration, and position constraint. To identify LOSAPs, a scenario-perception-assisted LOSAP search algorithm is introduced. It assumes that the Access Points (APs) installed in the current localization scenario are LOSAPs and that their measured distances represent line-of-sight (LOS) ranges. A Bayesian estimation-based indoor scenario perception model is constructed to assist in LOSAP identification, achieving a scenario recognition accuracy of 98.45%. In parallel, a calibration model based on Gaussian Process Regression (GPR) is developed to correct measured distances in LOS environments, improving ranging accuracy by at least 28.86%. To ensure that the estimated location remains within the current positioning scenario, scenario boundary information is utilized to constrain the positioning result derived from the corrected LOS distances. Experimental results demonstrate that the proposed method outperforms conventional approaches such as Least Squares (LS) and the Extended Kalman Filter (EKF), achieving an average localization accuracy of 0.875 m and a root-mean-square error (RMSE) of 1.014 m. Compared to LS and EKF, the proposed algorithm improves localization accuracy by at least 33.4% and reduces RMSE by at least 38.48%.
Keywords:
Smartphones
LOSAPs
Bayesian estimation
Gaussian process regression
Least squares
Journal
IF:
4.6
Papers:
1.5K
Citations:
1.6W