The WiFi Indoor Positioning System Based On Path Loss Model

With the advent of the ear of intelligence and the Internet of things,the location service is moving forward to people's daily lives as a new model.Currently,GPS and Beidou satellite navigation system has been able to achieve meter level positioning service in civil outdoor navigation.But it is not enough to achieve the indoor positioning that the weak signal can be captured in the building.However,most people spend more than half their time indoor,high accuracy and low cost indoor positioning services is very promising.The Wi Fi indoor positioning technology mainly uses the attenuation relationship between the signal and distance to calculate the location.Without additional hardware equipment,the technology can utilize APs arranged in large shopping mals,office buildings and other indoor environment to realize localization.But in the complex indoor environment,the propagation of the signal is affected by many factors,such as the indoor layout,personnel density,multi-path effect,which makes it difficult to build accurate signal propagation loss model.Therefore,the indoor positioning technology based on Wi Fi to achieve high accuracy is a practicality and challenging research topic.This paper mainly studies the Wi Fi indoor localization method based on the path loss model.Firstly,the principle of indoor positioning which mainly is on the two methods fingerprint matching algorithm and path loss model based on Wi Fi is analyzed and the differences between the two methods are pointed out.In order to better study the path loss model,the influence of the four factors including person,the different indoor partition materials,other AP and the height on the RSSI is analyzed by the method of control variables.In order to achieve high-precision positioning in the Wi Fi indoor positioning system,the construction of the path loss model and the positioning algorithm are the most important aspects to improve,which also is the breakthrough point in the paper.In this paper,the influence of different loss factors on the traditional path loss model is analyzed,and the results show that the traditional model is not suitable for complicated indoor environment.In the comparative analysis of the structure and properties of BP neural network and RBF neural network,select the RBF to build the path loss model,and introduces its principle and its construction method.And the simulation results show that the way to build a more appropriate description of the attenuation relationship between RSSI-Distance.There are three kinds of common positioning algorithms,which are the three sides and three triangle positioning,the least square method and the hyperbolic location method.In this paper,the principle of the three common algorithms is analyzed in detail.On the basis of the common algorithms and the theory of Taylor series expansion,an improved method is put forward to solve the problem of low positioning accuracy.This optimization method considers the two aspects that the fact that the closer the distance is,the bigger the RSSI and only one positioning is likely to lead to more big error.What's more,in the whole positioning process,the RSSI is sorted first,and then the coordinates are calculated by grouping to form a positioning loop.The centroid coordinate is taken as the final localization result,and its effectiveness and feasibility are verified by simulation.Finally,the paper realizes the Wi Fi indoor positioning system based on the path loss model under the Android platform,which is based on the Client/Server mode.The system makes the indoor electronic map through transforming the pixel coordinate into the WEV Mercator projection coordinate system.The data acquisition and map display are carried out on the client side,and the logical operation such as the data processing and the positioning algorithms is completed on the server side.This APP provides a real-time,relatively accurate positioning and displays the location of the point to be located on an electronic map.In order to make the APP function more diversified,a analog path planning and simulation navigation are added.