Design And Implementation Of GSM Network Positioning System Based On Location Cluster

With the development of geographical information system and mobile communication, the requirement of positioning people, vehicle and other mobile targets has been placed on the agenda. The prospect of location-based service is cheerful. In recent years, research of wireless location mainly focused on two aspects, one is designing efficient and reliable algorithm for location; the other is obtaining the accurate parameters of an existing algorithm in complex and varied wireless environment.This paper mainly focused on the fast positioning technology based on location cluster. Fast position technology is the supplement and enhancement of China Mobile’s existing location service platform. It provides self location for popular valued-added services which exist as mobile client software. Comparing to present technologies, fast positioning technology has more advantages such as quick response, precise location and self learning optimization. It can effectively promote users’ service experience. To getting high quality’s database for positioning and promoting response time and accuracy of positioning, the following improvements have been made based on fingerprint location technology in this thesis:1. The primitive data were processed according to setting period. Invalid and repeated data were wiped off; the valid were saved into database. Static data of base station were also updated base on the expiration.2. Clustering technology was introduced into mobile location. Radius and density of cluster and base station’s threshold, which was used to filter collected data, were computed by the base station’s coverage. Filtered data were clustered and divided by time.3. As clusters’shapes were different and irregular, smallest circle containing all points in a cluster is applied. Center and radius of the smallest circle were saved into database which was used by fast positioning platform. 4. Algorithm in container was chosen by reading algorithm configuration. The algorithm called data that handled with cluster and coverage of smallest circle to compute mobile station’s position. The center and radius were sent as position and error distance separately back to user. The information of the location was also saved to relative database.Prototype of fast positioning platform was realized in this thesis. Test of the algorithm was carried out. The result showed the algorithm can achieved96meters by67percent in urban and achieved1000meters by94percent in suburban. The system fulfilled the position request of Beijing mobile research institute basically and has been deployed in Beijing mobile and Liaoning mobile.