【作者】 杜巧玲；

【导师】 钱志鸿； 王树勋；

【作者基本信息】 吉林大学 ， 通信与信息系统， 2009， 博士

【摘要】 节点定位技术是无线传感器网络的重要研究内容之一。目前绝大多数算法是针对无线传感器的二维网络开展定位。但是实际上,无线传感器网络节点经常处于三维环境中,如森林、海洋等,因此需要提供节点的三维位置信息。由于无线信道的传输损变,使得节点定位的定位精度无法保证。而在网络配置之前,需要按预定的网络参数(如节点密度、锚节点密度、无线射程等)掌握该网络的定位精度。因此分析在一定条件下的定位精度是无线传感器网络的一项关键技术。本文以无线传感器的二维网络为基础,以提高定位精度为目的,从信号源模型、信道模型出发,对三维定位算法、定位精度以及定位方法等内容进行了深入研究。本文从影响无线传感器网络的定位精度的主要因素出发,分析信号源模型,选择合适的信道衰减模型。将无线传感器网络的二维节点定位算法的三边算法扩展到了三维网络中,提出了无线传感器的三维网络节点定位的最小二乘算法。为了进一步提高算法的稳定性,提出了三维网络节点定位的基于信噪比的加权最小二乘算法。采用Matlab仿真,仿真了不同参数对三维的三边算法、最小二乘算法和加权最小二乘算法的影响。分析结果表明,所提加权算法能够在获得最大覆盖率的条件下,在确保高稳定性的同时,具有较高的定位精度。本文对基于RSSI、TOA技术的无线传感器网络,以锚节点位置固定条件下的二维定位精度CRB分析为基础,提出了锚节点随机分布情况下的三维定位精度CRB分析。首先给出了RSSI、TOA技术下三维定位精度CRB的表达式,然后仿真分析了锚节点密度,无线射程对定位精度CRB的影响。仿真结果表明,三维位置估计的克拉美-罗下界能够取得和二维的克拉美-罗下界一致的效果,可以作为三维位置估计的性能的衡量标准。本文针对无线传感器网络定位技术,提出了一种定位锚节点的新方法。通过对磁偶极子模型的分析,利用二阶差分的方法,得出了在二维网络中,随机撒播锚节点的情况下获得锚节点的绝对位置的方法。采用matlab仿真,仿真实验分析了该定位方法的有效性。同时,该方法容易实现锚节点的替换工作。更多还原

【Abstract】 The node localization is one of the most important research aspects in wireless sensor networks. However, most of the previous research limits localization to a two-dimensional (2D) space and does not provide enough information to infer three-dimensional (3D) location. But in the context of real-world applications, large quantities of sensor nodes are a physical impossibility to be airdropped into the area of absolute plane forming a ground WSN. The physical world is 3D scenario, e.g., deep sea, hill or wild nature. Due to the wireless channel attenuation, the precision of the node localization can not be protected. But the localization precision should forestall the deployment of wireless sensor networks to be predominated under some network parameters (such as node density, anchor node density and radio range). So the analysis of node localization is also an important technology for wireless sensor networks.Many localization methods in WSN have been proposed by far. These methods are usually aimed at how to locate unknown nodes or how to improve performance of localization under a certain density of anchor nodes. However, localization method for anchor nodes is limited to GPS or hand-placing method merely. Perhaps the simplest method of providing localization is to equip every anchor node with a GPS receiver in outdoor environments]. However, GPS technology presented a costly solution for localization in WSN, due to cost, form factor, energy consumption, and the requirement for a second radio. Another alternative is hand-placing each anchor node and manually recording its position. This is a tedious and error prone approach unsuitable for large sensor networks and many of the proposed WSN applications. This creates a demand for efficient and cost-effective localization discovery methods of anchor nodes in WSN. Therefore, the research contents of our paper are developed based on these problems. The main achievements and innovations in this dissertation focus on the three aspects as follows. Firstly, the weighted least squares method of node localization. A weighted least squares method is presented by using the ambient noise. The weighted least squares methods is presented in RSSI scenarios. Simulation results indicate that the weighted least squares method has excellent localization precision.Secondly, based on 2D Cramer-Rao analysis on condition that anchor nodes was placed in fix position for wireless sensor networks, the corresponding 3D Cramer-Rao analysis on condition that anchor nodes was distributed randomly are proposed in this thesis. With a matlab simulation platform, CRB was simulated respectively in two application scenarios, RSSI and TOA. The influence of anchor node density and radio range on the CRB was analyzed. Finally, A new localization method of anchor nodes was been proposed in wireless sensor networks. The magnetic dipole model was analysed. The obtained absolute coordinates method of anchor nodes was provided in the case of placing anchor nodes randomly under 2D flat. The result of simulation indicates the performance of the localization method. And the method can resolve easily the problem that new anchor nodes replace old anchor nodes.The dissertation consists of seven chapters.In chapter one, the introduction and the analysis of the development and the present situations of wireless sensor network are summarized. The theoretical significance and the practical values of purpose and contents of this dissertation are given.Chapter two introduces the basic methods of node localization for wireless sensor networks. Sum up all sorts of benchmark about localization, and summarize the previous localization method.In chapter three, the node localization of MLE for 3D wireless sensor networks。The triangular algorithm of 2D is applied to 3D system, and the MLE arithmetic is presented in the case of 3D wireless sensor networks。In chapter four, the weighted least squares method of node localization. A weighted least squares method is presented by using the ambient noise. The weighted least square methods is presented in RSSI scenario. Simulation results indicate that the weighted least squares method has excellent localization precision. In chapter five, 3D Cramer-Rao analysis for wireless sensor networks. Based on 2D Cramer-Rao analysis on condition that anchor nodes was placed in fix position for wireless sensor networks, the corresponding 3D Cramer-Rao analysis on condition that anchor nodes was distributed randomly are proposed in this thesis. With a matlab simulation platform, CRB was simulated respectively in two application scenarios, RSSI and TOA. The influence of anchor node density and radio range on the CRB was analyzed.In chapter six, study on the node localization for wireless sensor networks under magnetic dipole model. A new localization method of anchor nodes was been proposed in wireless sensor networks. The magnetic dipole model was analysed. The obtained absolute coordinates method of anchor nodes was provided in the case of placing anchor nodes randomly under 2D flat. The result of simulation indicates the performance of the localization method. And the method can resolve easily the problem that new anchor nodes replace old anchor nodes.In chapter seven, a brief summary of the dissertation is given, and future work is also expected.更多还原