Wireless Sensor Networks Localization Method Under Field Conditions

Wireless sensor network (WSN) is widely used in these years for its advantages such as self organization, low cost and low energy consumption. Especially in the wild environment applications. This paper based on the field of cultural heritage protection and wildlife protection and aimd to solve the specific problems of WSN system in them. This paper suggests two different methods for the improvement of sensor localization and device-free target localization based on RSS dynamics under the condition of large random deployment in the wild environment.The specific research work includes the following two aspects:1) A sensor localization method based on region segmentation and regions cooperation under the condition of large random deployment in the wild environment.The energy consumption and economic costs of WSN sensor localization system becomes the most important two elements in wild environment, especially in the severe environment. DV-Hop algorithm has its advantages under this condition because of it’s low costs of hardwares and low power consumption. However, because of the existence of obstacles and non-uniform deployment, DV-Hop may cause unacceptable error in the practical application. This paper suggests a sensor localization method based on region segmentation and regions cooperation and solve these two problems with sensor connectivity information. The method has low cost and imporve the localization accuracy at the same time under different size of nodes and beacon ratio.2) The suggestion of localizability and its improvement of device-free target localization based on RSS dynamics under random deployment in the wild environmentThe device-free target localization based on RSS dynamics utilize the RSS dynamics caused by the targets to locate the position of them. Through the theoretical and experimental proof, we found that there are non-localizable regions under the condition of random deployment with this method and suggest the conception of localizability. And established localizable area coverage model. Then we hold a method to optimize the localizability via voronoi diagram and delaunay triangulation. Compared to the random method, the method in this paper needs fewer nodes in the same optimization goals and has better optimization effective with same number of nodes addition.