Research Of Indoor Positioning Technology Based On WLAN And Fingerprint

With the development of wireless network,the location-based service market is growing rapidly.In its technique supports,the wireless positioning technology has become the focus of people's attention.Credited to the wide deployment of Wireless Local Area Networks(WLAN)indoors,people can easily receive the wireless signals through their mobile phones,PDA and other mobile devices.Therefore,WLAN has created favorable conditions for the development of indoor positioning system.As WLAN and fingerprint based positioning system can readily utilize the existing equipment without changing the terminal's hardware.As it can cover a wide coverage are with low cost and high precision,WLAN and fingerprint based positioning system are widely used.Through in-depth analysis of WLAN and fingerprint based positioning system and with particular focus on characteristics of the fingerprint database,and the problems in similarity search and positioning accuracy,This paper put forward the corresponding respectively algorithms,and algorithm parameters were optimized,based on cluster analysis and the DS evidence theory.Firstly,through basic field experimentation,the paper analyzes the features,of the key signals that influence the formation of fingerprints database,such as Human impact,Rate of received signal,Distribution of signal strength.This provides experimental basis and practical guidance for the building of accurate fingerprint database.Secondly,building on basics of cluster analysis and combining the difficulties of slow processing the large amount of data in the similarity search positioning,this paper conducts Gaussian clustering analysis on Fingerprint database based on signal characteristics.On this basis,this paper proposes a fast similarity search algorithm,which greatly reduces the positioning calculation while speeding up the search in the calculation of the effective range.Finally,this paper,starting with information fusion among different sources as well as considering the uncertainties of same collection during the positioning phase within positioning algorithm,try to achieve the objective of improving positioning precision.Then,this paper,with DS evidence theory as the theoretical basis,treating each AP as a source of information,put forward multi-feature matching algorithm based on evidence theory.This algorithm is then simulated and the parameters of it are optimized.Through comparing with other relevant algorithms,and based on experimental analysis,this algorithm is proven to be effective and feasible.