Research On Indoor Positioning Technology Based On WIFI

With the increasing demand of users for positioning service,the traditional outdoor positioning technology can not meet the needs of consumers for indoor scenes,and the precision of outdoor GPS navigation positioning has reached meter level.Because of the complexity of indoor environment,it is difficult to locate GPS accurately.In recent years,the popularity of WLAN wireless network and the rise of smart phones,the powerful features of mobile phones provide consumers with rich and colorful experience,people gradually satisfied with the use of mobile phones to achieve life,entertainment,business and other activities.It can be said that mobile phones have become an essential part of people's lives.therefore,Using mobile phone as positioning tool and WIFI wireless network as carrier,WIFI indoor positioning technology has become a hot topic.WiFi indoor positioning technology has many advantages,such as simple operation,low cost and no additional equipment,etc.In this paper,the indoor positioning technology of WIFI is deeply studied.This paper starts with the research of WIFI indoor location technology,and through the analysis and comparison,it is found that the WIFI position fingerprint location method is more suitable as the indoor location technology method.However,the traditional location fingerprinting method still has some problems,such as low location accuracy and low location efficiency caused by the high complexity of matching algorithm.In the offline phase,due to the complexity of the indoor environment,the collected data usually contain a lot of noise.If the acquisition data is not processed properly,the generated fingerprint database will not match the current location environment.In the on-line stage,because there are too many reference points in the indoor area,it will be very complicated for online matching,and the traditional clustering is very important.The method can not match reasonable reference points for some special anchoring points.When calculating the similarity between fingerprint points,it is also necessary to consider whether the traditional Euclidean distance calculation method can obtain the appropriate reference points.In order to further improve the above problems,this paper analyzes and improves the two aspects of the off-line data preprocessing and on-line matching algorithm of the position fingerprint algorithm.On the basis of analyzing the source of data and pre-processing,a mean filtering algorithm is proposed by using the strongest AP filtering method and Gao Si filtering method,on the basis of the common preprocessing algorithms.Aiming at the well constructed fingerprint database,this paper analyzes the disadvantages of using the neighborhood mean filtering method of fingerprint database to deal with the noise signal,and proposes a neighborhood filtering technique based on the variance of AP signal.Traditional K-means clustering algorithm for matching Operation In this paper,a new algorithm based on clustering algorithm is proposed,which can improve the positioning accuracy of the locating points which exist in the boundary-like boundary.The algorithm is the same as that of the common locating points.The algorithm is based on the clustering algorithm,which can improve the localization accuracy of the locating points in the class boundary.Finally,based on the traditional Euclidean distance to calculate fingerprint similarity,a formula based on weighted Euclidean distance is proposed,and an algorithm based on weighted Euclidean distance is obtained.The accuracy problem of using traditional Euclidean distance matching reference points in positioning process is solved.