AQU-WiFiLoc: Context-Aware Indoor Localization Based on WiFi Fingerprinting
Date
2026-08-25
Authors
Ali Jamoos
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
This paper presents AQU-WiFiLoc, a context-aware indoor localization framework that augments Received Signal Strength Indicator (RSSI) based conventional Wi-Fi fingerprinting with Cisco Access Point (AP) telemetry and magnetic field signatures. A multi-modal dataset containing 1,343 fingerprints was collected from 14 rooms distributed across two floors of the Faculty of Engineering at Al-Quds University, Palestine. Initial experiments showed that all evaluated classifiers achieved perfect classification performance under clean conditions, indicating that the collected radio map provides highly discriminative room-level fingerprints. To evaluate robustness under more realistic operating conditions, additive Gaussian noise was introduced exclusively into test-set RSSI values while auxiliary contextual features remained unchanged. Four machine learning classifiers and six feature configurations were evaluated under progressively increasing RSSI degradation. Results demonstrate that contextual information significantly improves localization robustness. The proposed Context-Aware configuration achieved a macro F1-score of 98.5% using K-Nearest Neighbors under severe noise conditions with standard deviation of 10 dBm, outperforming the RSSI-only baseline. In addition, AP telemetry provided the strongest individual performance gain, highlighting the value of network-side context for robust indoor positioning.
Description
Keywords
Citation
M. Leqyanya and A. Jamoos, "AQU-WiFiLoc: Context-Aware Indoor Localization Based on WiFi Fingerprinting," 2026 IEEE Jordan Conference on Applied Electrical Engineering and Computing Technologies (AEECT), Amman, Jordan, 2026, pp. 485-490, doi: 10.1109/AEECT69724.2026.11657872.