Font Size: a A A

Indoor Positioning Algorithm For WLAN Based On KFCM-LMC-LSSVM

Posted on:2019-06-28Degree:MasterType:Thesis
Country:ChinaCandidate:Y WangFull Text:PDF
GTID:2428330548485895Subject:Electronic and communication engineering
Abstract/Summary:PDF Full Text Request
With the continuous development of wireless network technology and computer applications,Location-based Services(LBS)has been researched and applied in many areas of life,which is mainly used to facilitate people to get the required location information in real time.At present,the common means of outdoor location,Global Positioning System(GPS),is difficult to receive signals in the indoor environment and can not meet the requirements of indoor location service.Therefore,the positioning technology for indoor complex environment is created.Compared with various indoor positioning techniques,the received signal Strength Indication(RSS)based on Wireless Local Area Networks(WLAN)is not required by other hardware and equipment,and is widely used in real life.In indoor positioning environment based on WLAN,in order to obtain higher positioning accuracy,a lot of work must be done to collect the RSS value of a large number of fingerprint points,while the RSS signal is time-varying,and it is necessary to update the fingerprint library regularly.To solve this problem,the indoor location algorithm is studied theoretically and experimentally,and the location effect of several algorithms is analyzed and compared.In this thesis,the region is divided by kernel fuzzy C mean clustering.After judging the area of the point,the fingerprint library model is quickly established by using the low rank matrix filling theory in this area,and the model is solved by the inexact Lagrange multiplier algorithm,and the fingerprint library with high density fingerprint points is reconstructed.Finally,we use least squares support vector machine to train RSS values and physical coordinates of fingerprint points,and locate the physical location of the points to be measured.The experiment shows that the KFCM-LMC-LSSVM algorithm has high positioning precision,and can effectively reduce the workload of fingerprint data acquisition,and can well meet the requirements of WLAN indoor positioning.In the end,the Chinese Xuan paper culture park in Jingxian County,Anhui,is applied in the actual scene.According to the requirements of the project,the indoor positioning system based on the Android platform is realized.The system combines the location of the electronic label and the WLAN positioning to show the tourists the detailed information of the tourist paper cultural tourism,and provide the map navigation and positioning.Sights information push and other services.Through practice verification,the system can be positioned accurately and has certain practical value.
Keywords/Search Tags:Wireless local area networks, Indoor positioning, Fuzzy clustering of kernel functions, Low-rank Matrix Completion, Least Squares Support Vector Machine
PDF Full Text Request
Related items