Research Of The Positioning And Tracking Algorithm For Moving Target In Wifi Network

With the rapid development of data and multimedia business, the demand for location services is growing. No matter people in outdoor or indoor environment, the demand of quick and accurate position-obtained of mobile terminal and location service has become increasingly urgent. Location-service is playing an indispensable role in our life. At the same time, WiFi network communication protocol which is based on the IEEE 802.11 a/b/g/n was put forward. It has flexible topology, and improve the conditions of the existing wireless network communication. So it quickly became important supplement of the mobile communication system, and has become a worldwide research hotspot now.In this thesis, we improve the traditional algorithm of average cluster and kalman filter, and come up with the positioning technology for moving targets in WiFi network. The technology is mainly used for the environment which under the coverage of WiFi network. It can position and track the mobile terminal according to WiFi signal strength. Based on the analysis of the existing WIFI positioning technology and the problem of traditional K-means clustering and kalman filter algorithm, the algorithm of the positioning technology is improved as followed:1. Through the analysis of the fingerprint characteristics on the location coordinates sequence, we introduce clustering algorithm fingerprint processing. After the study of various clustering algorithm, we improve the K-Means clustering technology with the specific situation of the indoor WiFi positioning which in order to reduce the phenomenon that selection of the initial clustering center clustering results lead to the result unstable and inaccurate. On the other hand, in order to solve the heterogeneous type terminal positioning problem of poor effect, serialization comparison algorithm is introduced.2. With the analysis of position sequence of the targets from the positioning system, we first propose an adaptive algorithm based on kalman filter. It updates automatically by the error of observation data, leading to reducing the dependence on the scene environment in traditional kalman filter, and improving the project valuation. On the other hand, in order to solve the filtering divergence problem, we put forward the attenuation factor of adaptive, which can ensure not only the filter convergence but also reducing the old data influence on filtering result..3. This topic finally also implements fast positioning tracking system combining with the above technology, and do a test on the performance of the system. However, experimental results confirm the system can meet project requirements under the coverage of WiFi network. The contents of this paper are supported by the special fund of the MIIT for the Internet of things, the central university basic research fund projects and so on.