WiFi Indoor Location Technology Based On The Fingerprint

With the increasing number of the large indoor spaces in the modernized cities,such as large shopping malls,underground parking,hospitals,underground shopping districts and so on,more and more people stay here for activities.In these indoor spaces,people need to know their own location information,and then planning routes according their needs.Because of the influence of the wall and other shelters,people can not be positioned by GPS indoors,many domestic and foreign scientific research institutions began to develop indoor positioning technology.Indoor positioning technology is a new technology based on different signals such as UWB,Wi Fi,ultrasonic,laser and so on.According to the existing WiFi signal in the large indoor space,the indoor location based on WiFi signal can reduce the increase of the equipments.Meanwhile,the existing mobile terminal-mobile phone can be used as the positioning terminal,which opens up a new direction for the use of Wi Fi signal.This paper investigates the status of indoor positioning technology both in Chinese and foreign.Aiming at the problem of the existing indoor positioning based on WiFi fingerprint location method.A new kind of fingerprint library is proposed:MAC physical address classification and the signal intensity of the RSSI clustering built on the traditional fingerprint library.By using the attenuation model of WiFi signal in space,the region space of MAC classification is determined,and the errors caused by the jump of WiFi signal are solved.The regional fingerprint database is further subdivided into K clustering subsets by the introduction of K-mean clustering algorithm.The WiFi signal with similar signal intensity is classified as a clustering subset.In order to optimize the Wi Fi signal in the cluster subset,the noise signal in the clustering subsets is filtered by the neighborhood mean filtering method.The subdivided fingerprint database further reduces the error caused by the different location fingerprints when calculating the positioning distance.In this paper,the WKNN method is used to extract the coordinate information of neighboring fingerprint points and calculate the location coordinates by weighted calculation.Through the simulation analysis of different fingerprint databases,the validity of the clustering fingerprint database is proved to improve the location accuracy.At the end of this thesis,a real-time positioning system is built on Android platform.Through the experiments of real-time positioning of different Android mobile phones in the real environment,the validity of clustering fingerprintdatabase about indoor location clustering is tested.