Research On Key Technologies Of WIFI Location Fingerprint Indoor Positioning Algorithm

With pervasively available of wireless networks and rapid development of mobile smart devices,high-precision indoor location in healthcare,emergency rescue,social networking,navigation and surveillance field have demonstrated huge market prospects.WIFI-based location fingerprint algorithm is uncomplicated, low cost and can meet the accuracy requirements in most indoor positioning applications,becoming the focus of research for indoor positioning technology.WIFI-based fingerprinting positioning technology is studied and researched deeply in this work. On the basis of analysis of potential positioning error and its causes,the article explores four key areas contained in offline and online phases of fingerprinting positioning: fingerprint database construction,partition,AP selection and matching estimation,proposing the comprehensive improvement programme of WIFI-based location fingerprint indoor positioning algorithm to lower computation cost and optimize accuracy. Simulations verify obvious increase in positioning accuracy and running efficiency using inproved scheme.Subsequently, a partitioning method based on graph coloring theory is put forward to replace complex clustering algorithm according to spatial data similarity.The method constructs the adjacent relation graph of reference points,according to the received same number of AP.Vertex coloring algorithm can achieve better clustering effect and reduce the computation cost compared with traditional iterative clustering operation.Ultimately, the article raises a hybrid indoor positioning scheme combining with the advantanges of WiFi and Bluetooth,reducing computation workload and dependence on the reliability of fingerprint database. The hybrid algorithm gives the specific scheme of Bluetooth hot deployment and regional determination method,greatly ameliorating real-time requirement and positioning accuracy with low cost.