【作者】 冯砚毫；

【导师】 曾孝平； 江禹生；

【作者基本信息】 重庆大学 ， 通信与信息系统， 2011， 硕士

【摘要】 无线传感网节点定位技术是无线传感网重要的支撑技术之一。节点位置信息不仅是事件监测的前提,也是网络拓扑自配置、提高路由效率、向观察者报告所需信息的基础;不知道所获信息所对应的位置,所获信息也失去了应有的意义。无线传感网节点部署是随机的,由于受到布设环境和自身因素的影响,如何最大限度提高节点定位精度是无线传感网定位研究领域的一个难点。因此,开展节点定位技术的研究具有重要理论和现实意义。本论文介绍了无线传感网发展历程和结构模型,分析了无线传感网中的关键技术,深入讨论了无线传感网节点定位技术。着重研究了节点定位的概念、定位技术分类、定位的基本原理和基本方法,对节点定位的性能评价标准进行了较详细地分析研究,同时对现有的典型定位算法和定位系统进行了分析比较。论文重点对无线传感网定位技术中的非测距定位DV-hop算法进行了深入研究,详细分析了其存在误差的原因。分别在二维和三维环境下提出了改进的节点定位算法。(1)为了减小传统DV-hop算法的定位精度误差,论文在二维环境中提出了一种新的DV-hop定位算法。通过引入平均跳距权值、跳数权值和距离误差,改进未知节点平均跳距的计算方法,使估算出的距离能更接近实际距离;通过引入共线性阈值和跳数阈值,选择拓扑关系好且距离未知节点较近的锚节点组进行位置估计得出一系列位置;最后通过质心算法得出最终的位置坐标。仿真结果表明新算法在不需要任何额外硬件的支持下,能提供比传统DV-hop算法更精确的位置估计。(2)针对新DV-hop定位算法存在覆盖率较低的问题,提出了一种升级虚拟锚节点的定位算法。通过把一些达到条件的未知节点升级为锚节点的方法,对新算法进行进一步的改进,有效地解决了覆盖率的问题。仿真实验结果表明了该方法不仅能够有效地解决覆盖率问题,而且能在在一定程度上提高定位精度。(3)在三维环境下,提出了一种新的三维节点定位算法。采用DV-hop距离矢量跳数的思想和数学几何学原理,引入位置估算偏差值利用加权质心算法进行位置估算。仿真结果表明了该算法的可行性。更多还原

【Abstract】 The wireless sensor network node location technology is an important supporting technology for WSN. The node position information of WSN is not only the premise of monitoring events, but also it is the basis of the network topology from the self-configuration, improve routing efficiency and reporting the information required to observers. If we do not know the position information, the information we are interested will lose its proper meaning. Wireless sensor network nodes are randomly deployed. How to maximize the localization accuracy of localization is a difficult research area in wireless sensor networks for the effect of the layout environment and factors of its own. Therefore,the research for the wireless sensor network node location technology has important theoretical and practical significance.This paper introduced the development course of wireless sensor network and the architecture model of the wireless sensor network, analyzed the key technology for wireless sensor networks, and thoroughly discussed wireless sensor network node location technology. Focused on the concept of node localization、the location classification、the basic principles and methods of localization, and the criteria of positioning performance evaluation are made a more detailed analysis and research, while some typical existing location algorithms and systems are made a brief introduction and comparison. The paper focused on the Range-free positioning algorithm of wireless sensor networks DV-hop in-depth study, and analyzed the reasons for the existence of error in detail. Then, this paper presented the improved localization algorithms in two-dimensional and three-dimensional environment respectively.( 1 ) The paper presented a new DV-hop localization algorithm in the two-dimensional environment to improve localization accuracy error of traditional DV-hop algorithm. The average hop-distance weight、hop weight and distance error were introduced into the algorithm to improve the node average hop-distance, then the calculated distance through using the average hop-distance will be closer to real distance; And then the anchor node teams which have better topology and are closer to the unknown node are selected through co-linearity threshold and hop threshold to estimate location and get a series of locations. Finally, the final coordinates can be got through using Centroid algorithm. Simulation results show that the novel DV-hop algorithm can provide more accurate location estimation. (2)For the novel DV-hop localization algorithm have the problem of low coverage, a location algorithm which upgraded the virtual anchor nodes algorithm was presented. In order to solve the coverage problem, the paper made a further improvement for the novel algorithm by upgrading unknown nodes which can meet some requirements as anchor nodes. Simulation results show that the method can effectively solve the coverage problem, and can improve the location accuracy.( 3 ) In the three-dimensional environment, the paper presented a new three-dimensional localization algorithms by employing the distance vector hop ideas of DV-hop and principles of mathematical geometry and Centroid algorithm to estimate node position. Simulation results show the feasibility of the algorithm.更多还原