Research On DV-Hop Algorithm For Wireless Sensor Networks

In 1990s,with the rapid development of embedded systems, wireless communications, distributed information processing technology and microelectronic mechanical systems (MEMS) technologies, the wireless sensor network, which is constitutive of sensors with the ability of sensing, computing and wireless communicating, has aroused great concern. Deployment,coverage,nodes localization and network communication protocol are several basic questions of research on Wireless Sensor Network. Among these questions, nodes localization is one of the main supporting technologies. This is because many applications of wireless sensor networks depend on the location information of sensor nodes.Limited by volume, cost, power and hard to get close to the work place, the basic localization way is to deploy a few anchors, other sensors compute their own coordination by estimating distance, angel, location relation and network connectivity.The thesis provides a detailed introduction and complete analysis to the node localization of WSN. It analyzes and compares some typical localization algorithm such as RSSI, AOA, centroid location algorithm, APIT, DV-Hop, and the thesis focuses on the investigation of DV-Hop localization algorithm. By collecting the minimum number of hops between unknown nodes and anchor nodes, DV-Hop algorithm estimates the unknown node's position coordinates using hop distance instead of actual distance. As a Range-free algorithm, DV-Hop has advantages, but inevitably there are also some shortcomings. Based on DV-Hop algorithm,a improved node localization algorithm is designed for solving the above shortcomings. Improved algorithm finish the unknown node location after multiple rounds by limiting the number of hops from anchor nodes. And located nodes will be raised to anchor nodes involving the unknown nodes location. Then, the two algorithm are tested with Matlab platform. The experimental results show that the location accuracy and coverage of the improved algorithm are significantly improved compared with the DV-Hop algorithm. Finally, we sum up this thesis and prospect for future work.