Research Of3D Node Localization For Wireless Sensor Networks

Wireless sensor network is known as intelligent networks after computer and internet. It integrates features such as sensor technology, MEMS and wireless communication technology, which has "self-organization, self-healing" characteristics. If the internet provide an advanced platform for exchange for people, sensor network reflect a real objective world for people.This paper studies nodes positioning technology which is the key technique of wireless sensor networks, and focusing on the three-dimensional location. A number of localization algorithms had been put forward for sensor networks, but most is used in two-dimensional space. The three-dimensional space is more in line with the actual deployment environment of the node. This paper focuses on research methods of the three-dimensional positioning at this stage, proposing the following algorithm.As one of the classical algorithm for sensor networks, APIT algorithm has the advantage of simple, practical, highly efficient, with good accuracy and low communication overhead by range-free positioning method. It will be extended in three-dimensional space in the paper. Our major contributions are as follows: firstly introducing APIT location principles and methods, improving for two location error and constraints under sparse node environment, achieving simulation results show that the improved algorithm reduces positioning errors and is more suitable for different forms of network; Secondly, following the APIT ideas, replacing PIT test by space tetrahedron test, regarding space tetrahedron as different form by various grid for to filter obtained space location. Because the positioning method not only can be retained characteristic of range-free position method, but also can be applied to the three-dimensional positioning for a variety of scenes, it has advantage of high value.This paper proposes a three-dimensional space location algorithm which is based on the cell division. Our major contributions are as follows:firstly making a detailed introduction to the division of the grid, carrying on the analysis from three respects that include expectations of the location area, required beacon node number of positioning and cell density; secondly introducing centroid algorithm to improve positioning accuracy, using weighted iterative method to improve the localization of the coverage area. The simulation results show that, under fixed conditions of the beacon node, the algorithm of iterative weighted centroid of the cell division method positioning accuracy of space unit meshing method has an average of20%improved, the positioning algorithm without the complexity of distance measuring equipment and expensive external facilities will be able to get high positioning accuracy and positioning ratio, and relatively simple communication protocol, network topology has a certain robustness.Aimed at the shortcomings of low accuracy of the three-dimensional positioning, matching node algorithm which based on ranging technology is put forward. The core of this method is to complete the screening of the node with the spatial location of the distance constraints, and then use the matching node group to complete positioning. To provide a reference which is three-dimensional positioning mechanism for wireless sensor networks, the least squares method in the equation is used to solving phase position refinement, and meeting all kinds of positioning accuracy needs by selecting different matches.