Research On Fingerprint Construction And Matching Algorithm In WLAN Location

With the rapid development of wireless communication technology,the demand for the location-based services is increasing in our daily life.Location-based services have a wide range of demand and application in many areas,including mobile commerce,parcels tracking,and social networks.Positioning is the basis of locationbased services.It can be divided into indoor positioning and outdoor positioning.In the field of outdoor positioning,the Global Navigation Satellite System(GNSS)is usually used to locate the user.GNSS has relatively high accuracy in outdoor positioning.However,GNSS is not suitable for indoor positioning because of the obstruction of indoor buildings.Therefore,many researchers focus on indoor positioning technologies in recent years.Indoor positioning technologies based on bluetooth,infrared and ultrasonic usually require additional infrastructures,which limits its application in actual scenarios to some extent.With the development of wireless communication networks and the popularization of Wireless Local Area Networks(WLAN),the indoor positioning system based on WLAN is very convenient to deploy without additional hardware costs.Therefore,WLAN positioning has become a research hotspot in recent years.This thesis mainly focuses on the construction of fingerprint database in the offline phase and the matching algorithm in the online phase in WLAN positioning.The main research contents are as follows:(1)An automatic fingerprint database construction method based on region division is proposed.It aims at the problem of large labor cost when collecting fingerprint database in offline stage.In the actual indoor positioning environment,the positioning area is divided into several sub-areas according to different obstacles.By collecting several reference points,the environmental parameters of each sub-area are estimated using least squares parameter estimation.Finally,the offline stage fingerprint positioning database is automatically constructed according to the propagation model to avoid large-scale manual data collection.Simulation results show that the proposed algorithm has better performance.(2)Aiming at the problem of resource and time consumption caused by the demand of traversing the entire fingerprint database in the online phase,the database in the offline phase is clustered by Fuzzy C-Means(FCM).The fingerprint database is divided into several fingerprint clusters according to the difference of received signal strength indicator(RSSI).In the process of localization,we use the NN(Nearest Neighbor)matching algorithm to find the cluster which the target belongs to.Finally,considering that the traditional weighted K-Nearest Neighbor(WKNN)algorithm uses a stationary K value and leads to low accuracy,we use the dynamic WKNN algorithm to locate the user.