【作者】 张帆；

【导师】 吴信才；

【作者基本信息】 中国地质大学 ， 地图制图学与地理信息工程， 2014， 博士

【摘要】 随着以智能手机、笔记本电脑、平板电脑等为代表的智能移动设备的出现和发展,移动互联网得到了越来越广泛的应用,基于移动平台打造的工具软件和移动服务如雨后春笋般涌现,为各类行业用户和大众用户提供了丰富的服务。在移动互联网中,基于位置的服务是使用频率最高、应用最为广泛的服务之一,许多其它移动应用都直接或间接的使用到了LBS。借助于全球卫星定位系统,在户外等开放性场所,LBS已经能够为用户提供高精度、高稳定性的位置服务,其应用已渗入到各个行业领域和大众市场。然而在人类活动更多的室内场所,由于建筑物遮挡等问题,接收到的卫星信号往往已发生畸变,进而无法通过卫星定位系统准确地测量到目标在室内场所的准确位置。因此,研究适合室内环境的定位技术,开发满足行业用户和大众化应用需求的定位系统移动计算领域的研究热点之一。本文的研究工作是在近些年来无线局域网技术飞速发展和广泛普及的时代背景下,结合智能移动服务对定位导航技术、位置服务等新的需求而产生的。本文对这一基于时间测量的室内定位技术进行了较深入、系统的研究,并针对影响定位精度的关键因素加以归纳分析,并提出了相应的优化算法。通过与现有方法的比较分析,以及原型系统的实验验证,证明了本文的一系列优化算法具有一定的可行性。论文的主要研究内容和创新点归纳如下：(1)基于异步伪距差的室内定位方法的研究。本文提出了基于距离差的往返时间测量法,往返时间法(RTT)测量由目标终端主动发送的定位请求信号经过接入点处理返回的总的循环时间,以此避免对时钟同步的需求。同时采用伪距差法能够减少临近接入点因传播途中所受同源干扰造成的测量偏差。(2)往返时间算法中系统处理延时的实时测定。传统的系统延时测定往往采用预先测值法来求得,在以后的定位使用中,沿用预先测量到的系统处理延时。这种方法的特点是,初步测量的延时在以后的使用中保持不变,无法实时改变。本文所述的方法将在每次信号传递到接入点端时开始计时,在信号传递回给目标终端时将计时数据打包,同其它信息一起返回,终端根据收到的实时测定结果来计算位置,具有较高的定位精度。(3)以高斯“背景”模型的方式实现测量结果对动态环境的自适应。本文提出了一种“背景”建模的方法,解决测量数据变化后的自适应选取：将定位过程开始首段统计数据结果看作“背景”,“背景”为有效统计数据的合集。由于终端在室内场所的移动过程通常是缓慢的线性变化,当环境主动短时间变化,“背景”将保持不变,而当环境被动改变或长时间改变时,则将“新环境”(新的测试结果)线性的融入到“背景”中,形成新的“背景”数据,定位过程正是基于“背景”数据来测量终端的位置。(4)研究基于定位高程坐标来优化水平定位结果的方法。目标大多数情况下在室内的活动处于离散的平面空间内,高度值趋于稳定,往往在该楼层高度值的1-2米之内。有基于此,本文提出了一种基于高度误差来优化水平定位结果的方法,通过对多次测量结果的统计数据,予以评估,进一步剔除高度偏差较大的测量结果,以达到优化测试数据的目的。(5)研究以加权投票的方式实现定位信号接入点的实时选择。本文将研究一种加权投票方式来实现接入点的实时选取,其基本原理是选取信号传播途中,所受干扰相同的两个临近接入点,采用本文的距离差法计算两者信号传播的距离差,尽可能的保证信号传播途径中同源误差的去除。加权投票的方式主要用于接入点“临近”程度的判定。更多还原

【Abstract】 In the mobile Internet, the location-based service is the highest frequency of use, one of the most widely used service application, and many other mobile applications are used directly or indirectly to LBS. By means of a global satellite positioning system, such as in outdoor open spaces, LBS has been able to provide users with high-precis ion, high stability location services, its application has been infiltrated into various industries and the mass market. However, indoor places more activity in humans, due to the building block and so on; the received satellite signals are often distorted, and thus can not accurately measure the exact position of the target in the indoor place through a satellite positioning system. Therefore, the study environment for indoor positioning technology, the development of one of the focus areas of mobile computing industry positioning system users and popular application requirements are met. Research work of this paper is in IEEE802.11wireless local area networks in recent years, the rapid development of technology and the promotion of a wide range of backgrounds, combined with mobile computing needs new positioning technology, location-based services, such as on arising.In this paper, this time based on measured indoor positioning technology for a more in-depth study of the system, analyze and synthesize the key factors affecting the accuracy of positioning against, and the corresponding solutions. By comparison algorithm and experimental analysis proves the effectiveness and feasibility of the program. The main point of the paper work and innovation as follows:(1) Research asynchronous pseudorange differential based indoor positioning methods. In this paper, based on the distance difference between the round-trip time measurement, round trip time method (RTT) measurements to locate the target terminal sends a request to the total cycle time signal processing returns through the access point, thus avoiding the need for clock synchronization. While using pseudorange differential method can reduce the spread of the way near the access point due to interference caused by homologous suffered measurement bias.(2) Round-trip time measured based on delay system processing algorithms. Conventional systems often use a pre-determined time delay method to obtain the measured value, after the positioning of use; the system follows the process previously measured time delay. Feature of this method is that the delay in the preliminary measurement remains unchanged after use can not be changed in real time. The methods described herein will pass signals at each end of the start time to the access point; the signal is passed back to the timing data package to the destination terminal, returned along with other information, the terminal based on the real-time measurement results received to calculate the position, having a high positioning accuracy.(3) In this paper, the Gaussian "background" way to measure the results of the model to achieve adaptive dynamic environment. This paper proposes a "background" modeling approach to solve the adaptive changes in the measured data selection:The process begins with the first paragraph of positioning statistics results as a "backdrop",’background" for the collection of valid statistics. Since the mobile terminal in the indoor premises of the process is usually slow linear change when the environment changes the active short,"background" will remain unchanged, and when the environment changes or prolonged passive changed, the’New Setting"(New test results) into a linear "background", the formation of a new "background" data, the positioning process is based on the "background" data to measure the position of the terminal.(4) This article will coordinate research-based approach to optimize the positioning height horizontal positioning results. Activities indoors under most circumstances are within the target discrete flat space, height values stabilized, often highly value the floor within1-2meters. There Based on this, we propose a horizontal positioning to optimize the results based on height error method, the results of multiple measurements statistics evaluated further removed a large height deviation measurement results in order to achieve the purpose of optimizing the test data.(5) This paper will examine ways to achieve a weighted voting positioning signal in real time to select the access point. This article will examine ways to achieve a real-time weighted voting to select the access point, the basic principle is to select the signal propagation on the way, the interference suffered the same two adjacent access points, using the method of this paper is calculated from the difference between the signal propagation distance difference, as far as possible to ensure removal of signal propagation pathway homologous errors. The main method of weighted voting is used to determine the access point "near" level.更多还原