Algorithm Research And Positioning System Design Of Indoor WIFI Location Fingerprint

With the explosive growth of Internet of Things and mobile Internet applications,a large number of applications have increased the demand for Indoor Location Based Service(ILBS).Through the use of indoor positioning system can accurately obtain the user's location information,analyze its potential behavior,and recommend its potential needs.At present,the WiFi network has covered most areas of people's lives,including schools,shopping malls,airports,hospitals,and office buildings.Moreover,the WiFi network has the advantages of high transmission rate,low cost,and wide coverage,and its inexpensive receiver module has also been widely embedded including sensors,smart phones,home appliances and other devices,making use of existing WiFi network resources to achieve location fingerprinting has great application market and commercial prospects.Therefore,the WiFi location fingerprinting technology based on Received Signal Strength Indication(RSSI)become the most popular indoor positioning technology.It has application value in the research of positioning algorithm of WiFi location fingerprints and the design of practical positioning systems in specific scenarios.In this paper,two improved location matching algorithms are proposed for the indoor location area mainly based on " aisle " scenes and AP distribution.They are the fusion algorithm and the weighted Bayesian algorithm respectively.The fusion algorithm is a fusion of Bayesian algorithm and WKNN algorithm based on the strongest AP method;the weighted Bayesian algorithm is an improvement of the Bayesian algorithm based on the strongest AP method.Both improved algorithms are first coarsely localized to a smaller area by the strongest AP method and then accurately positioned by an improved location matching algorithm.The experimental verification of the positioning algorithm is performed in different actual scenarios.The results show that the positioning performance of the two improved algorithms in different scenarios is improved.In the scenario where the point to be positioned coincides with the fingerprint point,the fusion algorithm achieves a positioning accuracy of 82% within 3.6 m accuracy and the weighted Bayesian algorithm is 76%;in the scenario where the point to be positioned and the fingerprint point do not coincide,the weighted Bayesian algorithm has a positioning accuracy of 100% within 3m accuracy and the fusion algorithm is 96%.In both scenarios,the positioning performance of the two improved algorithms is better than other classical algorithms.Based on the algorithm research,this paper designs and implements an indoor location system based on WiFi location fingerprint on Android platform.The system software uses a hybrid architecture.The client includes the data collection side,positioning and navigation side,and the server is used to store the database,positioning and navigation algorithms.The positioning system has been tested and perfected.It has been successfully applied in the building of Information Institute,Yunnan University,and it has good scalability.