Comparative Analysis of Trilateration and Fingerprinting for Wi-Fi-Based Human Localization

This study compares two Wi-Fi-based human localization techniques—trilateration and Wi-Fi fingerprinting—within a controlled indoor environment using three strategically placed Wi-Fi access points (APs). Trilateration estimates user positions by calculating distances from the APs based on Received Signal Strength Indicator (RSSI) values, while Wi-Fi fingerprinting matches real-time RSSI readings to a pre-constructed database using the K-Nearest Neighbors (KNN) algorithm. Conducted in a university hostel bedroom, the experiments evaluated the algorithms across five positions. Trilateration faced limitations from signal attenuation and multipath effects, achieving correct localization in only 3 out of 5 positions. In contrast, Wi-Fi fingerprinting matches real-time RSSI readings to a pre-constructed database using the K-Nearest Neighbors (KNN) algorithm, yielding superior results by correctly identifying 4 out of 5 positions. Optimized with k = 3, Wi-Fi fingerprinting achieved 76.50% overall accuracy and 75.30% validation accuracy. These findings demonstrate Wi-Fi fingerprinting as a more reliable method for indoor localization in complex environments.