Effective Fingerprint Extraction and Positioning Method Based on Crowdsourcing

At present, WiFi fingerprinting indoor positioning technology is a research hotspot, and the construction of a radio map based on crowdsourcing can significantly reduce the amount of labor required. However, when users collect fingerprint information using crowdsourcing data in practical applications, the received signal strength (RSS) information collected by user phones updates slowly and will remain the same for a certain distance and time. Therefore, crowdsourcing data is inaccurate, and a radio map thereby established will also lead to inaccurate indoor positioning. In order to address this issue, this paper proposes a method for extracting effective RSS information from crowdsourcing data in order to establish the radio map and achieve positioning. By analyzing the crowdsourcing data characteristics, effective RSS information is identified and extracted. Based on inertial sensors data, the position of effective RSS information is determined. The effective crowdsourcing data associated with the position is weighted and fused in order to establish the radio map. Aimed at the issue of the RSS collected by phones updating slowly, a combined positioning method for position calculation is proposed in order to achieve effective positioning. Experiments were conducted in a real environment to validate the algorithm, and the positioning accuracy could reach 1.5 m by using crowdsourcing data to construct the radio map and executing an indoor positioning algorithm, which is close to the accuracy achieved when using manual data collected by professionals at a certain distance and saves a great deal of labor.