Hybrid Indoor Positioning System for Pedestrians With Swinging Arms Based on Smartphone IMU and RSSI of BLE

Abstract

Over the last decades, researchers have invented many positioning algorithms for pedestrians in the indoor environment, such as trilateration, fingerprint, and pedestrian dead reckoning (PDR) system. However, indoor positioning is still difficult to solve with current technology. The fingerprint and trilateration based on received signal strength indicator (RSSI) usually require filters to stabilize the RSSI signal, which will cause a delay in positioning, which is fatal for real-time positioning. The PDR system with the inertial measurement unit (IMU) of smartphones usually requires the direction of the Y-axis of smartphones parallel with movement. Nevertheless, pedestrians will swing their arms while walking normally, which is not considered in the traditional PDR. In this case, a huge bias in heading estimation is inevitable, which causes the traditional PDR system to be unavailable during walking with swinging arms. In this article, a hybrid indoor positioning algorithm based on IMU and RSSI is proposed for pedestrians with swinging arms. The PDR system is improved by analyzing the characteristics of walking postures. In order to eliminate the cumulative error existing in the traditional PDR system, we propose the multipoint positioning algorithm based on RSSI for calibration. Combined with the Kalman filter, positioning delay and error have been effectively reduced for indoor location-based services (ILBSs).