西安科技大学 硕士学位论文 无线传感器网络定位技术的研究 姓名：张新亮 申请学位级别：硕士 专业：计算机应用技术 指导教师：龚尚福；郭秀才 2011-06 论文题目：无线传感器网络定位技术的研究 专 业：计算机应用技术 硕 士 生：张新亮赵 静 (签名) 指导老师：龚尚福 (签名) 秀才 郭秀才 (签名) 摘 要 近年来，随着微机电系统(Micro-Electro-Mechanical Systems)、无线通信技术、传感 器技术与集成电路技术的迅速发展， 无线传感器网络已经成为当今前沿性的热点研究方 向之一。 将无线传感器网络定位技术应用于井下监测系统， 通过传感器节点协作的感知、 采集和处理，实现对网络覆盖区域内事件的探测、识别、定位和跟踪，不但能够提高井 下监测系统的性能，而且检测系统能在事故发生后继续运行，为搜救工作提供信息。 通过对无线传感器网络定位技术的研究，结合井下定位环境的特点，论文选择无线 传感器网络中无需测距的定位技术作为研究方向， 研究无线传感器网络中节点的自定位 方法。 (1)对于静态节点定位方面， 论文研究了基于多跳的定位机制， 重点研究了 DV-Hop 节点定位算法,该算法通过少数锚节点的动态刷新使多数待定位节点实施定位。 通过分析 DV-Hop 算法中距离估计误差大、定位精度差等不足，结合井下监测网络拓扑的非规则 性特点，论文对算法定位过程中节点间距离估计和定位输出这两个阶段进行优化，给出 了基于 DV-Hop 节点定位算法的新方案：DV-HopZ；(2)对于动态节点的定位方面，针对 锚节点动态刷新方法表现出实时性差，精度不高且能耗较高的问题，论文研究了蒙特卡 罗移动定位算法。为了提高算法的实际效用性，论文对定位过程中预测、过滤等核心环 节进行改进和优化，并且引入预测分析中的插值理论，给出了一种基于动态节点局部采 样、 实时输出的新方案： MCLZ； (3)在对静态节点和动态节点自定位方法研究的基础上， 使用 MATLAB 仿真工具，设计并实现了基于 DV-HopZ 和 MCLZ 两种方案的节点定位 原型系统。原型系统能够仿真显示两种方案的定位性能，网络中节点的部署情况，网络 节点的连通性情况。 仿真实验表明，论文中对于静态节点和动态节点定位的新方案，能够满足井下监测 系统的定位需要，具有良好的定位精度，较少的定位能量消耗和较强的网络适用性。 关 键 字：无线传感器网络；节点定位；DV-hop 算法；蒙特卡罗定位算法 研究类型：应用研究 Subject : Study on Localization Technology in Wireless Sensor Network Specialty : Computer Application Technology Name : Zhang Xinliang (Signature) Instructor : Gong ShangfuLi (Signature) Guo Xiucai (Signature) ABSTRACT Recent years, with the rapid development of Micro-Electro-Mechanical Systems, the wireless communication technology, the sensor technology and the integrated circuit technology, the wireless sensor network has already become one of leading edge hot spot research directions. Apply the wireless sensor network positioning technology in underground monitoring system. Through the sensor node collaborative sensation, collection and processing, realizing to the network coverage of the region events detection, the identification, the localization and the track, which can enhance improve the performance of the underground monitoring system, moreover, the detection system can run to continue after the accident, provide information for search and rescue work. Based on wireless sensor network positioning technology research, combining the characteristics of underground positioning environment this paper take wireless sensor networks without ranging positioning technology as a research direction to research wireless sensor network the node from the localization method. (1) For the static node localization aspect, the paper has studied based on the localization mechanism, which focuses on studying the DV-Hop node localization algorithm, the proposed algorithm through a few anchor nodes that most dynamic refresh backlog a node implementation positioning. Through the analysis of the DV-Hop algorithm mid-range estimation errors, poor location accuracy insufficient, combining the underground monitoring network topology non-regularity characteristics, The paper on the rules of node during localization algorithm is estimated and the distance between the two stages positioning output the optimization, putting forward the new plan named by DV-HopZ of node localization algorithm based on DV-Hop .(2) For dynamic node localization aspects, in view of the anchor node dynamic refresh method , showing bad real-time, precision and high energy consumption, the paper studies the Monte Carols mobile location algorithm. In order to enhance the actual utility of the algorithm, the paper making the improvement and the optimization in the process to forecast and filtrate the core links, and in the introduction of predictive interpolation theory, proposing one kind based on the dynamic node partial sampling, the real-time output new plan: MCLZ; (3) for the static node and the dynamic node from the localization method studys foundation, using the MATLAB simulation tool, designed and realized two kind of plan node localization prototype systems based on DV-HopZ and the MCLZ. The prototype system can demonstrate the localization performance of two kinds of plans, in the network the node location state and the node connective situation. Simulation experiments show that the paper for static nodes and dynamic node positioning of the new plan, which can satisfy the need of underground monitoring system, with good location positioning accuracy, few positioning energy consumption and strong network applicability. KeyWords: Wireless sensor networks Node localization DV-hop algorithm Monte Carlo localization algorithm Thesis : Application Research 1 绪论 1 绪论绪论 1.1 课题研究的背景和意义 近年来，随着微机电系统、无线通信技术、传感器技术与集成电路等技术的迅速发 展，传感器网络已经成为当今前沿性的热点研究方向之一，其应用有着巨大的科学意义 和应用前景。传感器网络是由大量传感器节点所构成的新型信息获取系统，每个传感器 节点具有有限的感知、计算和通信能力，通过专用网络协议对信息进行协作的感知、采 集和处理，从而实现对网络覆盖区域内发生的事件进行探测、识别、定位和跟踪1-3。 传感器网络的低成本、可以随意分布的特点，使传感器网络应用在了很多领域，特别是 在恶劣的作业环境领域。 对煤矿井下环境和人员活动信息进行监测，一直是一项热门研究课题。目前绝大多 数的矿用监测设备都采用有线连接的传感器采集并传输的方式来完成， 由于环境特殊性 使这种传统的有线布设方式存在着本质的缺陷：首先布