Research On Indoor Positioning System Based On Wi-Fi Fingerprint

Recent years have witnessed a spurt of progress in mobile intelligent device and the Internet of Things technology.People's demand for the the Location-aware applications has increased sharply.As a part of mobile Internet applications,location services have a profound impact on all aspects of people's travelling.In recent years,with the maturing of outdoor positioning technology,the overall net interest rates decline year by year,while the demand of indoor positioning continues to grow,so huge business value drives people to pay more attention to the development of the indoor positioning technology.Among them,the Wi-Fi indoor positioning technology become the most popular one with the advantage of wide range of infrastructure and high popularity of device equipment.Therefore,this paper will study the algorithm theory and system implementation of indoor positioning based on Wi-Fi fingerprint,and aims at the problem of device heterogeneous and low positioning accuracy in the application of positioning system in real environment,the ELM algorithm is used to calibrate the difference of each positioning device,so that the positioning device can obtain the same RSSI data as the training device.On this basis,the mapping matching technology is integrated to improve the positioning accuracy of the positioning device.The specific research contents are as follows:(1)Aiming at the device heterogeneous problem,this paper first analyzes the device hardware heterogeneous and RSSI signal distribution to study the causes of device heterogeneous.Then,a differential calibration algorithm based on ELM is proposed to reduce or eliminate the difference of the RSSI collected by each positioning device and improve the compatibility of Wi-Fi fingerprint positioning system to different positioning device.At last,the validity of the algorithm is proved by experimental comparison.(2)Aiming at the positioning accuracy problem,We study the model of Wi-Fi fingerprint positioning algorithm based on the calibration of positioning devices,and propose a positioning algorithm based on Wi-Fi fingerprint fusion map matching by fusing indoor electronic map information to do the spatial constraints of position estimation.What's more,map matching technology can be used to correct the Wi-Fi anchor point deviation,so as to reduce the positioning error,improve the positioning precision of the positioning system.Experimental results prove that,in proportion to the traditional positioning algorithm,this algorithm can validity enhance the positioning precision of the positioning system.(3)Combined with the research results of device heterogeneous calibration algorithm and location algorithm based on Wi-Fi fingerprint and map matching,this paper builds a positioning system on the Web platform,the functional requirements of the positioning system are firstly analyzed.Then,according to the functional architecture design,the fingerprint data collection function of the mobile terminal is designed and implemented,as well as the positioning service,end-user query service and historical track query service of the web client.Finally,the development and test of the positioning system are completed.