Research On Indoor WiFi Fingerprint Map Based On Crowdsourcing

With the popularity of wireless devices and mobile terminals,the research and development in the field of positioning has been rapidly promoted.Due to the continuous construction of large-scale modern buildings,people need to know their position information at any time in indoor activities,so the demand for indoor positioning is also increased.Among them,WiFi positioning technology has become the mainstream research direction of indoor positioning with its advantages of low cost and easy layout.However,the efficiency of constructing WiFi fingerprint map based on traditional collection method is low,and it takes a lot of time and labor cost.In order to solve this problem,this paper introduces the concept of crowdsourcing and proposes a lightweight solution to construct WiFi fingerprint map by crowdsourcing collection.In this scheme,the accurate PDR trajectory is obtained by using sensor data and map matching based Pedestrian Dead Reckoning(PDR).The mapping relationship between actual position and fingerprint information is completed to generate crowdsourced fingerprints.Secondly,considering the attenuation of the wall to the signal,it was added to the signal propagation model to conduct modeling analysis on the Access Point(AP)in the indoor area,and the theoretical value of the signal propagation model was combined with the actual measured value to complete the optimization of the original fingerprint database.Then,in order to solve the noise problem caused by too many AP data,the AP subset selection algorithm based on maximum information entropy was adopted to eliminate the irrelevant or redundant data,improve the fingerprint distinguishing ability,and realize dimensionality reduction of high-dimensional fingerprint data.Then,in order to solve the problem that fingerprint map may have uneven distribution,Kriging interpolation method is used to conduct fingerprint interpolation for areas with little or no fingerprints,so as to construct a more complete and accurate fingerprint map.Finally,based on the linear difference of Received Signal Strength(RSSI)among different terminal devices,an algorithm based on RSSI difference sequence is adopted to solve the problem of equipment difference caused by crowdsourced acquisition.In view of the scheme proposed in this paper,through the experiment of the scheme of each part of the content of research and analysis,to verify its effectiveness and reliability.In addition,by comparing with traditional acquisition way to construct WiFi fingerprint map,found that the scheme is not only in the localization accuracy have been improved,and greatly shortens the time of construction of fingerprint map build up to the light type of WiFi fingerprinting map.