Research On The Localization Method Based On RSSI For Wirless Sensor Networks

In recent years, with the rapid development of Wireless Sensor Networks(WSN). Localization become one of the research foundation of WSN. Location information is the necessary prerequisite to realize the application of Wireless Sensor Networks such as target identification, surveillance and tracking. Furthermore, Wireless Sensor Networks Localization is helpful to some applications such as routing schemes, load balance, topology control. So the study for Localization of Wireless Sensor Networks gets great significance in both theory and practice.After analyzed the status of Localization of the Wireless Sensor Networks sufficiently, the distance measuring method based on RSSI of range-based method with fine feasibility is used for researching based on the practice need of WSN, then the localization algorithm with high precision and self-update algorithm is studied later.Firstly, we use CC2530 to establish the hardware platform and ZigBee protocol to realize wireless communication. We combine MATCOM and Visual C++ to program the integrated control and data analysis software platform, therefore the real RSSI is collected and analyzed in different environments and technical support is provided to realize localization algorithm research and evaluation. We proposed the step-log-normal-model and fixed point clustering method based on the basic method log-normal-model and curve fit method to provide better distance measuring precision for the localization algorithm research. We compare and analyze the basic localization algorithms using real RSSI data and the Bounding-box algorithm with good precision has been chosen as the fundamental research algorithm. The weighted Bounding-box and circle-tangent Bounding-box algorithm are proposed aim at the large error region. The conclusion is good both in simulation and practice condition.In the end, this Dissertation studies the self-update algorithm to avoid prior information. And on the basis of the probability density self-update algorithm we proposed the Least Square self-update algorithm. Simulation shows that this algorithm converges fast with high precision.