The Technology of Crowd-Sourcing Landmarks-Assisted Smartphone in Indoor Localization

Abstract

In recent years, with the popularity of smartphones, mobile sensor technologies have been widely used in indoor localization. Its advantage is that data can be collected directly from smartphones without installing any other equipment. Meanwhile, it has the disadvantages of high requirements on positioning conditions and large deviations in positioning results. A technology of crowd-sourcing landmarks-assisted smartphone in indoor localization (CLS-IL) is proposed in this paper so as to solve these disadvantages. Firstly, the accelerometer, magnetometer and gyroscope sensors are used to acquire initial positioning of users. The characteristics of magnetometer and gyroscope sensors are analyzed through a series of experiments in allusion to the key problem of heading estimation errors, then corresponding data is matched for different experimental conditions. After that, the corresponding landmark database is established for experimental environment and an adaptive optimization algorithm of landmarks is proposed to correct drift errors caused by the readings of sensors influenced by surrounding environment and long positioning time excessively during traditional pedestrian dead reckoning. Experimental results show that the location accuracy of CLS-IL system is significantly improved compared with traditional technologies, which provides strong support for subsequent studies.

Keywords

• Adaptive optimization
• crowd-sourcing landmarks
• indoor localization
• mobile sensors
• smartphones