BROWSE
Detail
Conference Paper
Indoor Localization of Unknown Wi-Fi Access Points with Directional Bias Compensation
Cited 0 time in 
Share 
Share
- Authors
- Hansol Park, Youngjin Lee, Jaejun Yoo
- Issue Date
- 2025-10
- Citation
- International Conference on Information and Communication Technology Convergence (ICTC) 2025, pp.1742-1747
- Publisher
- IEEE
- Language
- English
- Type
- Conference Paper
- Abstract
- This study proposes a joint optimization pipeline that combines an environment-adaptive path-loss model with directional bias compensation to estimate the locations of unknown Wi-Fi access points (APs) in indoor environments. The pipeline (i) calibrates the path-loss exponent n, reference power R0, and directional bias bdir using anchor-AP data; (ii) estimates initial (x, y, R0) for each unknown AP; and (iii) jointly re-estimates all unknown AP coordinates together with n and bdir. Using RSSI collected at 48 sites (four headings) in the ETRI lab, we vary the data-usage ratio from 10% to 100% via group-based stratified sampling and evaluate performance using the 2D Euclidean distance between estimated and ground-truth coordinates. The overall mean localization error was 4.20 m, with per-AP mean errors of 1.00 m (AP_X), 7.44 m (AP_Y), and 4.15 m (AP_Z). The lowest average error of 3.23 m was achieved at 70% data usage, representing a 28% and 26% reduction compared to 10% and 50% usages, respectively. These results highlight that balanced spatial and heading coverage is more critical than just the number of samples. The proposed method robustly estimates unknown AP locations with limited data, demonstrating 70% data usage as an effective trade-off between accuracy and data cost.
- Keyword
- Unknown AP Localization, Wi-Fi Localization, RSSI, path-loss model, Directional Bias Compensation, Joint Optimization, Least Squares Optimization, Model-Based Positioning, Rogue AP Detection
- KSP Keywords
- Access point, Adaptive path, Average error, Bias compensation, Directional bias, Euclidean Distance, IEEE 802.11(Wi-Fi), Indoor environment, Indoor localization, Joint optimization, Least Squares Optimization