Research On WLAN Indoor Localization Based On Location Fingerprint

With the development of mobile intelligence group perception and the development of Internet of things technology,location information has become an indispensable basic factor.People are more and more concerned with the convenience of location informatio n.More and more commercial interests brought by indoor positioning services have promoted the development of indoor positioning technology.Indoor positioning has been in the past several decades.It is an active research field.Outdoor environment usually uses the global positioning system(GPS)to complete location determination.Because of the lack of sight of the satellite in the indoor scene,the effect of GPS positioning is poor,and several technologies are used to build efficient indoor positioning system.When choosing the best system,consideration should be given to the trade-off between cost,accuracy and time deployment.Because of the many advantages of WLAN,most indoor positioning methods use WLAN as the basic mode of location.WIFI based location fingerprint indoor location method has been paid more and more attention in the academic and industrial circles,mainly due to its popularity as wireless access technology and the popularization of WiFi network infrastructure and terminal.The impro vement of positioning accuracy and the updating of fingerprint database have become hot topics.In this paper,a new indoor location method based on kernel principal component analysis(KPCA)and least squares support vector regression(LSSVM)regression algorithm is proposed.As RSSI(received signal strength indication,received signal intensity)is affected by noise and multipath effect in the indoor WLAN environment,the correlation between the position fingerprint and the physical location is reduced.In this paper,the missing value processing method is used to preprocess the collected fingerprint samples and remove the abnormal values.Then,the KPCA method is used to further process the fingerprint library to extract the main features needed.Finally,the least squares support vector machine is used for regression modeling to construct the nonlinear projection between the position finger and the actual physical location.Shoot relation,and use genetic algorithm(GA)to optimize model parameters.The simulation results show that compared with the traditional location method,the proposed method has higher positioning accuracy and faster location speed.At the same time,this paper also studies the updating of fingerprint database with the change of indoor dynamic environment.With the passage of time,the change of AP position,the movement of people,and the switch of doors and windows,it directly affects the established fingerprint library.The established fingerprint library has not been well reacting to the current position.Therefore,it is one of the problems that must be solved based on the fingerprint location method to adapt to the environment change.In this paper,the orthogonal matching tracking algorithm is used to reconstruct the fingerprint library,update the fingerprint library,and finally use the Compressive sensing(CS)method to realize the positioning method.This method can achieve high positioning precision by a small amount of RSS measurement.