节点文献

基于事件驱动的节点定位方法研究

Research on Event-Driven Node Localization

【作者】 孙伟

【导师】 金乃高;

【作者基本信息】 大连理工大学 , 软件工程(专业学位), 2014, 硕士

【摘要】 无线传感器网络是一个由随机部署的传感器节点构成的自组织网络系统。正因为其自身的随机性,加之传感器节点受低成本、低能耗以及有限的作用范围等诸多因素的限制,无法获知所有传感器节点的位置。所以由已知传感器节点估计未知传感器节点位置的问题依然是一个十分重要的研究课题。高精度的节点定位方法对后续的目标定位方法奠定了技术基础。目前,基于测距和距离无关的定位方法是无线传感器网络节点定位中两类不同的定位技术。基于测距的定位方法一般有四类解决方案,包括测量无线电信号强度、信号传播时长、信号接收角度和信号干扰的方法。而距离无关的定位方法是有效地利用锚节点相邻性、网络连通性信息,直至最新提出的基于事件驱动的节点定位方法。基于多序列的节点定位方法是一种典型的基于事件驱动定位技术。本文提出的基于凸优化的事件驱动定位方法就是在此基础上完成的。该方法进一步挖掘节点序列中隐含的扫描顺序与节点位置的关系,将定位问题转化为用线性规划方法求解多个线性不等式最优解的问题。本文通过严谨的理论证明和大量的仿真实验来验证基于凸优化的事件驱动定位方法的定位性能。仿真实验的结果表明,与现有的节点序列方法对比,无论是随着锚节点数量变化、目标节点数量变化,还是随着扫描次数的变化,我们的方法均取得了较好的定位效果。特别是资源受限的情况下,如锚节点数量较少时,该方法依然可以达到很高的定位精度。这表明了该算法具有更好的实用性和更广的应用领域。

【Abstract】 Wireless Sensor Network (WSN) is composed of random distribution sensor nodes, and it is a self-organization network through radio communication. Considering the randomness and severe constraints on low cost, low power, and effective range of sensor nodes, most of sensor locations are not acquirable. Therefore, sensor node localization that using known node estimates unknown node is still a challenging task. High-accuracy node localization method can contribute to the research of target node localization.Most of the previous methods for WSN can be categorized into range-based localization and range-free localization. Range-based localization generally has four kinds of solutions, including signal strength, time of fly, angle of arrival, and radio interferometry. Range-free localization makes full use of anchor proximity, network connectivity, and then extends to the newest event-driven localization method. The multi-sequence positioning method is a typical solution of event-driven localization, thus this article proposes a Convex Optimization-based Method for Event-Driven Localization (COMEDL). COMEDL takes full advantage of the hidden messages from the relationship between node sequences and node locations. The sensor node localization problem becomes a convex optimization problem of solving a series of multiple linear inequalities.The performance of this design is proved by strict theoretical analysis and evaluated by extensive simulations. The simulation results demonstrate that with the increasing of number of anchors, number of target nodes and scanning times, COMEDL has improved accuracy than MSP localization method. Especially, when there is a limited resource, such as a small number of anchors, COMEDL can obtain better localization accuracy compared to other algorithm. As a result, our method has a better practice and more wide application.

  • 【分类号】TP212.9;TN929.5
  • 【被引频次】1
  • 【下载频次】28
节点文献中: 

本文链接的文献网络图示:

本文的引文网络