Indoor Positioning And Navigation Based On Mobile Devices

With the development of wireless internet and the rapid growth of urban modernization drive,the location based services(LBS)show giant potential in fields such as medical field,electronic commerce,emergency rescue and artificial intelligence,which puts the issue under the spotlight.And indoor positioning and navigation technology with high precision is the key point to achieve the LBS.The traditional Global Positioning System(GPS)and cellular mobile communication technology have quite high positioning accuracy in the open air.However,the accuracy decreases substantially under the influence of multipath effect and obstacles of the building.WiFi positioning technology based on Received Signal Strength(RSS)stands out among all these methods after an overall consideration about expected cost,positioning precision,follow-up maintenance,transport speed and good portability.It can realize indoor positioning on any mobile devices with a WiFi module through the full use of existing WiFi facilities without deploying any other hardware devices.However,RSS is susceptible to external environment interference,which seriously affects the stability and accuracy of indoor positioning system and makes the RSS-based positioning method to be far from the required precision for indoor applications.In response to these problems,this paper analyzes the characteristics of RSS and proposes an indoor location algorithm based on information fusion,which combines the estimated results of PDR and RSS-based localization by Kalman filter.And developing an indoor positioning,navigation and tracking application system on a smart mobile terminal in the end of the paper.The main contents and innovations of this paper as follows:(1)The construction of three-dimensional building space model.The building space model with clear structure,good expression ability and visualization effect is the basis of realizing indoor LBS.Compared to the outdoor environment,the complexity of the indoor space structure poses a great challenge to the indoor modeling.Based on the existing indoor data files,this paper designs and constructs a three-dimensional indoor space network model based on the "node-arc" structure to express the spatial attributes and topological structure of the indoor spatial elements.The principle of middle axis extraction in building corridor based on Voronoi diagram is discussed in detail,and the automatic extraction of single layer path improves the efficiency of modeling.(2)Improved RSS-based positioning method.Through the in-depth study of RSS,the paper analyzes the impact of different factors on the RSS in the view of indoor positioning.For the complexity and variability of RSS,a WKNN(Weighted K-Nearest Neighbor)indoor positioning algorithm with spatial convergence is put forward to realize the indoor positioning in a higher accuracy.At the same time,different access points(AP)selection and matching mechanism is used to remove the redundant AP data and optimize the AP positioning sub-set,which aims at improving the positioning algorithm efficiency and accuracy.Compared with other equivalent algorithms,the algorithm proposed improves positioning accuracy as well as its efficiency.In the experimental environment,the location fingerprint database was created at a sampling interval of 1.5m,and the average positioning error was 1.38 m when positioning AP number is 6.(3)Real-time tracking and navigation based on Kalman filter.In the real-time indoor navigation,since the RSS-based localization algorithm is susceptible to the changes of indoor environment,the positioning is unstable and with low accuracy.There is also an irregular jump in the continuous positioning of a moving object.Although the PDR algorithm can get a relative position prediction by the sensors of mobile device,the results' accumulative error cannot be eliminated.In this paper,the Kalman filter is used for data fusion and trajectory smoothing to get a higher dynamic positioning accuracy in the indoor navigation.And the filter system is reset at the corner to reduce the cumulative positioning error of the linear motion model at the turn.At the same time,the barometer is used in the navigation to identify users' downstairs behavior and finally developed a multi-floor positioning and navigation system on mobile devices.The experimental results show that the average error of the system is 1.2m during the indoor dynamic tracking and navigation and is a better stability in the cumulative error over time when compared to the PDR and WiFi localization results.