Tightly Coupled Bluetooth Enhanced GNSS/PDR System for Pedestrian Navigation in Dense Urban Environments

The performance of the global navigation satellite systems (GNSSs) can be degraded significantly in urban areas and semi-outdoor environments due to the blockage and reflection of GNSS signals. To enhance smartphone positioning in these areas, the integration of pedestrian dead reckoning (PDR) and Bluetooth low energy (BLE) sensors has been proposed and widely developed. However, the current loosely coupled (LC) integration algorithm often fails to achieve accurate positioning in complex urban environments. Challenges such as reliable assessment of GNSS quality, accurate estimation of pedestrian walking direction, and error due to variability of BLE signal (BLES) strength across smartphones and accumulated by PDR remain to be addressed. To achieve precise and continuous location service in urban areas, we have explored the characteristics of BLESs and integrated BLE with GNSS and PDR, enhancing positioning performance through three key algorithms. First, GNSS quality is monitored with the aid of BLESs and PDR results adaptively to minimize their impact on the integration results. Subsequently, the smartphone's orientation and the heading precisely estimated from BLESs are jointly utilized to determine the pedestrian's walking direction. Finally, a tightly coupled (TC) Kalman filter, incorporating distance measurements from strong-signal BLE transmitters, is employed to tackle the device heterogeneity inherent in smartphone BLE sensors and to reduce the cumulative errors associated with PDR positioning. Extensive experiments on different smartphones demonstrate that the proposed system consistently achieves positioning accuracy within 4 m in both urban and semi-outdoor environments.