Research On The TOF-based Localization Algorithm In WSN

Wireless sensor network is a network which is composed of a large-scale self-organizing sensor nodes. As an emerging technology, wireless sensor network has been widely used in military, industrial, agricultural, commercial, environmental science and other fields, which has a huge impact to human life. In many applications of wireless sensor networks, the location information of sensor nodes is crucial. Only getting their own positions in WSN, sensor nodes can quickly and accurately sense information in the target area and then process the relevant data. Therefore, the positioning technology of wireless sensor network is essential in all applications. In complex environments, NLOS often has great impact on performance of localization.This paper carried on the thorough research to NLOS error mitigation in the complex environment and localization algorithm in the networks with holes and then proposed some algorithms.The main work of the paper as follows:1.The paper summarizes several typical ranging methods and localization algorithms. It also analysis their basic principles, as well as points out their advantages and disadvantages with applicable scene. These provide theoretical basis for the study of localization algorithm based on TOF ranging in wireless sensor networks.2. For the NLOS errors contained in TOF ranging in complex environments, the paper proposes an algorithm named colored noise adaptive Kalman filter (ACN-KF)which can effectively eliminate the NLOS errors. In order to establish model of NLOS error in complex environment, the paper carried out indoor TOF ranging experiments which confirmed the NLOS error is not Gaussian, but colored noise. Colored noise is assimilated into the Kalman filtering process and the filter parameters is changed with the colored noise. At last, the algorithm get the real-time optimal estimation. Experimental results show that the algorithm can significantly eliminate the NLOS error of TOF distance and improve the range accuracy.3. For the traditional localization algorithms with low positioning accuracy in some networks with holes and poor communication, the paper proposes a localization algorithm based on ACN-KF. Firstly, two kinds of secondary nodes will be selected in the network which help to flood in the entire network and create trees. Find the notch nodes and the shortest paths between nodes in the network according to the trees. Then neighbor nodes carry out TOF ranging and define the path curved index for calculating the shortest distance. Finally the distance matrix is created according to these shortest distances. The iterative Association-function is used to solve the matrix and then obtain locations of nodes. Experimental results show that the algorithm significantly improves the performance.In summary, the paper focuses on the research on the performance of localization algorithm based on TOF ranging. And then two algorithms are proposed, which can eliminate NLOS error of TOF ranging and improve positioning accuracy. Simulation show the algorithms are effective and the results show the algorithm can effectively improve the positioning accuracy.