Improving WiFi Fingerprint Positioning Through Smartphone Built-In Sensors Based Trajectory Estimation.

Wireless indoor localization is fundamental to many smartphone applications, appealing to a great deal of research efforts in the past decades. Among them, RSSI based fingerprinting using existing WiFi infrastructures has become an increasingly popular technique. However, WiFi fingerprinting still suffers from the vulnerable and changeable wireless signal transmissions. To address this issue, we propose a novel smartphone built-in sensors assisted WiFi fingerprinting method. The data generated from the smartphone built-in sensors such as accelerometer and gyroscope is utilized to estimate the user’s trajectory. Then an integrated probability model combining the RSSI fingerprint and users’ trajectory is established to implement fingerprint-location match. Performance evaluation results show that the proposed method has higher location accuracy and better adaption to changeable environment.