无线传感器网络高性能定位算法研究

发布时间：2018-01-15 09:07

本文关键词：无线传感器网络高性能定位算法研究　出处：《大连理工大学》2014年博士论文　论文类型：学位论文

【摘要】：近些年来,随着无线技术的快速发展和传感器软、硬件技术的成熟,无线传感器网络(WSNs, Wireless Sensor Networks)引起了越来越多的国内外研究学者的关注。WSNs是一种新型的信息获取技术,因为其自身所具有的自组织性、动态性和数据为中心等众多特点,在军事和民用领域具有广阔的应用前景,例如,在工作人员无法到达的地震灾区进行人员搜救、灾后环境监控和余震检测。其中,数据为中心的特点决定了在其上进行的任何算法设计、应用开发都需要基于可靠而且高效的节点自身或者从周围自然环境中获取的各种类型的数据。节点的位置信息作为节点自身所具有的一种主要数据,对于很多基于位置的应用和服务,是相当重要的。本文针对高性能数据收集问题和利用高性能数据收集进行高精度定位问题,使用理论分析和实验评估相结合的方法路线,研究了无线传感器网络中数据高吞吐量、公平收集问题,提出了比例公平的高性能数据收集算法；研究了各种动态网络参数对定位精度的影响,发现节点的定位精度与定位计算时使用的数据(基于各种动态网络参数的不同取值,收集得到的用于定位的数据并不相同)之间存在关系。通过优化各种动态网络参数,提高数据收集的性能和所收集数据的针对性,提出了高精度的定位算法。主要工作概括如下： (1)在无线传感器网络中,网络数据收集的效率和数据收集的公平性问题是两个重要的研究课题,本文通过分析数据收集中“漏斗效应”问题以及挖掘对立问题：数据收集效率和数据收集公平性之间的相互关系,提出了一种高性能的比例公平的数据收集算法,在提高数据收集吞吐量的同时,达到了数据比例公平收集的目的。 (2)本文对周期性睡眠调度网络中的定位问题进行了详细的分析、研究,利用优化动态网络参数的方法,提高定位数据收集的针对性和有效性,提出一种高性能的周期性参数优化的动态定位算法。通过与当前已有的高精度定位算法的比较,发现利用参数优化确实能提高数据收集的针对性和有效性,从而提高定位算法的定位精确度。 (3)本文对基于移动数据收集的定位问题进行研究。通过深入分析各种移动参数对移动数据协助的定位算法的影响,以无线测距为手段,根据移动节点的观察模型,将节点的定位问题转化为基于某一移动节点位置坐标数据的后验概率坐标估计问题来处理。根据与其它几种经典的或最新的定位算法的比较实验和结果分析,本文提出的定位算法在多种参数配置下,在动态环境中,表现出了很好的定位性能,是一种高性能的定位算法。 (4)基于移动参数对定位性能影响的分析,本文提出了一种高性能的基于先验数据收集的室内定位算法LuPI,此算法利用了易于获得的RSS (Received Signal Strength)信息,以各个采样点之间的RSS差值作为先验信息,构建先验信息数据库,然后利用RSS差值数据库对移动节点进行定位,获得移动节点的相对位置坐标。本文利用移动智能设备和WiFi路由器实现了LuPI算法,搭建了原型系统,进行了真实环境的实验,通过与最新的LiFS算法的比较,LuPI确实提高了动态环境中移动节点的定位精度,并且能在环境复杂、多径干扰强烈的室内进行高精度的定位。
[Abstract]:In recent years, with the rapid development of wireless technology and soft sensor, hardware technology, wireless sensor network (WSNs Wireless, Sensor Networks) attracted more and more researchers focus on.WSNs is a new technology of information acquisition, because of its self-organization, dynamic and data the center and many other features, has broad application prospect, in military and civil fields such as earthquake can not arrive in the staff of the search and rescue personnel, environmental monitoring and detection of aftershocks after the earthquake. Among them, any number of algorithm design according to the characteristics of as the center of the decision on the need, application development based on node reliability but efficient or acquired from its surrounding natural environment in various types of data. One of the main data of the location information of the nodes as the node itself has, for it Location based applications and services are very important.
For the high performance data collection and collection of high precision positioning using high performance data line method using theoretical analysis and experimental evaluation of combining the research data in wireless sensor networks with high throughput, fairness is proposed for high performance data collection, proportional fair collection algorithm was studied; various parameters of dynamic network the positioning accuracy, using to calculate the positioning accuracy and the positioning of the nodes when the data (different values of various parameters, dynamic network based on the collected data for positioning is not the same). The relationship between the various parameters through optimizing the dynamic network, improve the performance of data collection and the data collected of the proposed localization algorithm high precision. The main works are summarized as follows:
(1) in wireless sensor networks, the fairness of network data collection efficiency and data collection are two important research topic, this paper through the analysis of the data collected in the "funnel effect" and explore the opposite problem: the relationship between data collection efficiency and fairness of data collection, presents a high performance the fair share of the data collection method in improving data collection throughput at the same time, to achieve the data proportional fair collection purposes.
(2) the positioning problem of periodic sleep scheduling in the network are analyzed in detail. The study, using the method of dynamic optimization of network parameters, improve the pertinence and effectiveness of positioning data collection, this paper proposed a dynamic positioning periodic parameter optimization of high performance algorithm. By comparing the algorithm with high precision positioning the current findings, pertinence and effectiveness of using parameter optimization can improve the data collection, so as to improve the positioning accuracy of the positioning algorithm.
(3) this paper studies the location problem of mobile data collection based on impact localization algorithm for mobile data to assist through in-depth analysis of various parameters to the mobile, wireless ranging means, according to the observation model of the mobile node, the node localization problem is transformed into a mobile node position coordinate data of the posterior probability estimation of coordinates based on the problem to deal with. According to the classical and several other or new positioning algorithm comparison experiment and result analysis, the proposed localization algorithm in a variety of configurations, in a dynamic environment, showing good positioning performance, is a localization algorithm with high performance.
(4) analysis of the influence of parameters on the performance of mobile location based on the proposed LuPI indoor positioning algorithm based on a priori data collection of high performance, this algorithm makes use of the easy access to RSS (Received Signal Strength), with each RSS difference between sampling points as a priori information, and then construct a priori information database. The localization of mobile nodes using RSS difference database, relative position coordinates access to mobile nodes. Using smart mobile devices and WiFi routers to realize LuPI algorithm, built a prototype system of real experimental environment, through comparing with the new LiFS algorithm, LuPI can improve the accuracy of dynamic positioning of mobile nodes in the environment, and in the complex environment, strong multipath interference for indoor high accuracy position.

【学位授予单位】：大连理工大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：TN929.5;TP212.9

【参考文献】