Indoor BLE and MEMS based multi-floor fusion positioning algorithm

Abstract

Based on the data fusion from micro electro mechanical system (MEMS) sensors and low-power bluetooth (BLE),an indoor BLE and MEMS based multi-floor positioning algorithm was proposed.First of all,the affinity propagation clustering,outlier detection and received signal strength indicator (RSSI) filtering algorithms were applied to denoise the fingerprint database.Second,by using the extended Kalman filter,the robust M estimation algorithm was used to perform the optimal estimation of the two-dimensional target position.Finally,the barometer output and geographical position information was considered to realize the height estimation of the target.The experimental results show that the proposed system is able to achieve the horizontal and vertical positioning errors lower than 0.7 m and 0.35 m respectively in multi-floor fusion positioning.

Keywords

• indoor positioning;multi-floor
• data fusion;BLE;MEMS