The Research Of Localization Algorithm In Underwater Wireless Sensor Network

With the development of land resources, the sea which is rich in resources for the future of social economy has played a decisive role. Under the vigorous development of all countries in the world, the proportion of marine economy to the world economy also presents an increasing trend year by year. The underwater sensor network has gradually become a hot research topic for many scholars. As development of land wireless sensor network, underwater sensor network have the characteristics of with wireless sensor network. For example, the energy of the node itself is limited and will be limited by the network bandwidth resources. And the particularity of the underwater environment makes the sensor network have the characteristics of large propagation delay, multi path effect and node mobility, which makes the development of underwater sensor network to overcome more difficulties. Underwater positioning technology is the key technology of underwater sensor network application in the face of various underwater applications. The data information that all application need should regard node location information as supporting, or it will get the data without any meaning. How to improve the positioning accuracy of the node becomes the key problem of underwater research work.In the current status of research in underwater positioning, positioning accuracy and node energy consumption is still the problem of research. Based on the analysis of existing underwater localization algorithm, we analyze error effect on the performance of localization and factors that influence the error. On the basis of least squares algorithm, we propose on the constraints of the weighted least squares algorithm of TDOA based on hierarchy. The algorithm firstly layer around the beacon node using the power control, then make beacon nodes with hierarchy and depth of information of unknown nodes lift the plane, which will be converted from three-dimensional positioning to two-dimensional localization. In ignoring the position error and range error is increasing, positioning performance of the proposed algorithm is superior to least squares algorithm. When the ranging error is fixed and the position errors exists, the algorithm can be achieved Cramer Rao lower limit at first, but whenσ2>2.5m2 performance divergence. Therefore, the algorithm is suitable for positioning error is small. Simulation results show that the algorithm of positioning accuracy compared with the least square positioning have been greatlyimproved.Based on the semi periodic model of seawater flow in the dynamic network of the coastal area, this thesis proposes a non time synchronization algorithm based on the sea water flow model. The algorithm firstly determines the location of the transmitter when the unknown node is requested. Then the non time synchronization algorithm is used to estimate the distance between the nodes. Finally, the position of the node is estimated based on the two order cone programming algorithm. Through the analysis of the number of beacon nodes, the initial clock offset and the clock offset of the node, it can be concluded that the proposed algorithm has a good performance in the comprehensive performance of energy consumption and positioning accuracy.The thesis has made a series of studies of the range-based localization algorithm.While considering the complexity of underwater environment, we should do further exploration and research on the accuracy of localization, node energy consumption as well as the dynamic change of sound velocity estimation.