Research On Localization Technology Of Wireless Sensor Networks

Since many applications of wireless sensor networks (WSN) depend on the location information of sensor nodes, how to obtain the accurate localization has become the focus of attentions. Without nodes' position, the gathered information is of little worth. So, the research on location technology of WSN has both theoretical and practical significance.After various localization algorithms are briefly reviewed. This dissertation mainly studied on the localization technology of WSN for indoor environment according to the characteristics and requirements of the location application for indoor environment, such as the peripheral areas of underground garage, museum and gymnasiums. Major contributions of this dissertation are as follows:(1) An micro-embedded real-time operation system for sensor nodes is designed. The hardware platform is a basic foundation of WSN research. Considering the characteristics of the sensor nodes named GAINS and the needs of research work, an embedded real-time operation software system which runs in the sensor nodes is designed by using the GNU C Compiler (GCC) language. It can work with limited hardware resource, organize and manage effectively hardware resources, support intensive concurrent operation and provide service for application.(2) A location algorithm based on RSSI is given and the experiments are carried out in the practical environment. First, we study on the characteristics of RSSI in real environments, then a conception of node's RSSI-Similarity degree in space environment is presented, which is used for weight factor to improve the weighted centroid localization (WCL). The indoor location experiments are carried out with no interferences and with interferences from the obstacle. The experimental results show that the improved algorithm has a smaller average location error and stronger robustness of environmental disturbance.(3) A location algorithm for moving nodes is proposed and has been experimentally studied in this paper. Since the position of sensor nodes may move in practical application, we study on the location technology of the mobile sensors in WSN. A new location algorithm base on Monte Carlo is proposed by limiting the range of samples and the RSSI value. It can solve the deployment of sensors and effectively decrease the computational complexity. To apply this algorithm, we established a WSN in the corridor of the building, using a robot which carried the sensor node to move in this network. The experimental results show that this improved approach has higher positioning accuracy.