Research On Range-Free Localization Algorithm In Wireless Sensor Network

Following the pace of rapid development that information age get forward, wireless sensor network has been widely used in military, medical, internet of things, environmental monitoring and other fields with the advantages such as low power consumption, low cost, distributed and highly autonomy. In this way, it becomes an interaction bridge between the human society and the physical world. Usually, sensor nodes are randomly deployed in a monitor area with a large scale, from the view of cost saving, only part of the node locations are known, which seriously affects the performance in applications of the wireless sensor network. As a consequence, node location technology, which is the basis for its application has get more and more attention. Based on distance or not, the localization algorithm can be divided into two categories. Aiming at the shortcoming of low positioning accuracy of the range-free localization algorithm when compared with the other one, in order to reduce the positioning error, optimize the positioning performance, this paper studies two of the range-free location algorithms, the Bounding Box algorithm and DV-Hop algorithm, and proposes corresponding improvement program.The paper achieved the following results:1) Overall first, this paper introduces the concept of wireless sensor networks and architecture as well as related basic knowledge. The typical positioning algorithms of wireless sensor networks are discussed, at the same time, the advantages and shortcomings of the various algorithms are briefly analyzed.2) Researched heavily on the Bounding Box algorithm, based on neighbor beacon nodes, using PIT judgment rule to find the best triangle, and further get the intersection with the original backlog area. In this way, it can narrow the scope that unknown node may exist, so as to improve the positioning accuracy. Experiment analysis show that the improved algorithm significantly improves the performance of the original algorithm.Based on the Bounding Box-PIT algorithm, after the object nodes located, it can use the information of unknown nodes which are in the range of one hop to narrow the target area again, the algorithm flow is given and the experiment analysis show that twice localization method based on the improved algorithm improves the accuracy of the algorithm.3) Against the accumulated error from every stage of the original DV-Hop algorithm, this paper proposes an optimization algorithm which is combined with genetic algorithm and tabu search algorithm to adjust the position of the unknown nodes got from DV-Hop. In addition, the algorithm flow is given and the comparison simulation is applied on the location error of improved algorithm through Matlab, the simulation results show that the improved algorithm has better positioning performance.