Research On Node Localization Algorithms For Wireless Sensor Network

As one of the state's five new strategic industry, the Internet of things has been received widespread concern in domestic. As one of the important basic research work of the Internet of things, Wireless Sensor Network (WSN) will continue to be a hot spot of research in academic circles. WSN is a distributed, multi-hop network composed of massive micro-sensor placed densely in the monitoring area in wireless, self-organizing manner, aiming to collaboratively sense, collect and process the information from the sensed objects in the monitoring area, and then report to the observer. Currently, WSN has played a great role in many applications such as military and national defense, health care, environmental monitoring, smart home, etc.WSN often distributes hundreds of sensor nodes into the corresponding application scene, so as to achieve specific monitoring information. Because sensor node is often distributed by random, its position cannot be determined beforehand. For most applications of wireless sensor networks, the monitoring information without location information is meaningless. In addition, the node's location information can be used to complete tracking, routing, balancing the network load and so on. However, the manual deployment of nodes would be limited by money, power consumption, scalability and other issues. Even, it might not be realized in certain situations. So, how to know the absolute position or the relative position of unknown nodes by the anchors is one of the most important issues in WSN.This thesis studies the existing classic localization algorithms, especially the centroid algorithm, the existing weighted centroid algorithms. Then, the thesis puts forward two new centroid algorithms based on power control.The Weighted Centroid Localization Algorithm Based on Power Control and Cooperative Theory makes full use of the power control theory to distinguish the anchors at different position by adjusting the anchors' transmit power. And every anchor will be set the reasonable weight for the every specific unknown node. In the situation of low nodes connectivity, the algorithm chooses some unknown nodes met the corresponding conditions to be auxiliary anchors, so as to reduce estimation error and promote the localization proportion.A Novel Weighted Centroid Localization Algorithm Based on Power Control uses the thoughts of power control to allot unknown nodes in the virtual grid by anchor's graded broadcasting. The algorithm calculates the vertex of the virtual grid and set the appropriate weighted for every vertex. Then, it calculates the weighted centroid of the vertex as the estimated location of unknown node. In addition, the position of unknown nodes can be further correction by properly setting the number of grading radio. The novel algorithm greatly improves the accuracy of positioning.