基于典型泛在无线信号的局域定位技术研究

发布时间：2019-01-17 14:27

【摘要】：随着信息时代的飞速发展和日益增长的生活出行需求,人们对自身位置信息的掌握也愈发急切,对位置服务也提出更高的要求。位置服务质量的好坏不仅取决于定位的准确性,还取决于位置服务区域的全面性,但社会城市化的发展却在制约着常规定位技术如GNSS的应用,无法满足人们在社会建筑群中的定位需求。在这种背景下,基于典型泛在无线信号WiFi的定位方式步入人们的视野,WLAN定位无需额外布设专用的导航定位设备,依靠现有周边的基础无线网络设备就能实现定位。但WLAN定位在实际场景中的应用,仍存在信号干扰波动性、定位算法完善性、WLAN接收设备差异性等问题,本文针对上述问题进行了以下研究工作。1、依据WLAN信号的传播特征,设计了两组室内实验,通过实验分析对比WLAN无线信号的RSSI时变特性和RSSI与距离关系,探讨不同信号强度下的信号波动特征和多路径效应对RSSI采样的影响。在其它滤波算法此基础上,提出一种临界高斯滤波算法,该算法弥补了单一的RSSI预处理方法的不足。2、利用Windows网络体系中的网络驱动程序接口规范NDIS,通过其中的Miniport Driver小端口网卡驱动,实现了WLAN无线信号RSSI提取;设计并创建了指纹库和指纹库索引,实现了位置指纹库的特征量的提取;引入欧氏距离权值解算定位点的坐标,实现了多维欧氏距离加权的WKNN算法的匹配定位,构建了一套完整的基于典型泛在无线信号的位置指纹库定位方法。3、完成了实际场景定位测试,测试内容包括静态单点定位、动态运动轨迹定位及不同无线网卡定位测试。静态单点定位探讨了不同K值选取下的点位定位误差,为局域定位的边缘区域定位提供参考;动态定位以轨迹线路的方式展现了WLAN定位下位置的连续变化情况;针对不同无线网卡在同一位置指纹库中的定位测试,验证了位置指纹库定位算法在不同接收终端互换条件下静态定位的可行性。测试结果表明,研究设计的基于典型泛在无线信号的局域定位方法在实际场景中定位切实可行,能实现室内环境下的静动态定位,并能取得较高的定位精度,基本能满足大部分室内环境下的位置服务需求,可以给相关需求的定位部门提供参考。
[Abstract]:With the rapid development of the information age and the increasing life travel demand, people are increasingly eager to master their own location information, and put forward higher requirements for location services. The quality of location service depends not only on the accuracy of location, but also on the comprehensiveness of location service area. However, the development of social urbanization is restricting the application of conventional positioning technology such as GNSS. Can not meet the needs of people in the social building group positioning. In this context, based on the typical ubiquitous wireless signal WiFi positioning approach into the field of vision, WLAN positioning does not need additional special navigation and positioning equipment, relying on the existing peripheral basic wireless network equipment to achieve positioning. However, there are still some problems in the application of WLAN localization in the actual scene, such as signal interference fluctuation, localization algorithm perfection, WLAN receiving equipment difference, etc. In this paper, the following research work has been done in view of the above problems. 1, according to the propagation characteristics of WLAN signal, Two groups of indoor experiments were designed. The RSSI time-varying characteristics of WLAN wireless signals and the relationship between RSSI and distance were analyzed and compared, and the effects of signal fluctuation characteristics and multipath effects on RSSI sampling were discussed under different signal intensities. On the basis of other filtering algorithms, a critical Gao Si filter algorithm is proposed, which makes up for the deficiency of single RSSI pretreatment method. 2. NDIS, is standardized by network driver interface in Windows network architecture. Through the Miniport Driver small port network card driver, the WLAN wireless signal RSSI extraction is realized. The fingerprint database and the fingerprint database index are designed and created, and the feature quantity extraction of the location fingerprint database is realized. This paper introduces the coordinates of Euclidean distance weights to calculate the location points, realizes the matching location of multi-dimensional Euclidean distance weighted WKNN algorithm, and constructs a set of complete location fingerprint database localization methods based on typical ubiquitous wireless signals. The actual scene localization test is completed, which includes static single point positioning, dynamic motion locus positioning and different wireless network card localization tests. Static single point positioning discusses the location error under different K values, which provides a reference for the edge location of local location, and dynamic positioning shows the continuous change of position under WLAN location by the way of track line. Aiming at the location test of different wireless network cards in the same location fingerprint database, the feasibility of static location of location fingerprint database location algorithm under different receiving terminal exchange conditions is verified. The test results show that the local localization method based on typical ubiquitous wireless signals is feasible in the actual scene, can achieve static and dynamic positioning in indoor environment, and can achieve high positioning accuracy. It can meet the needs of location service in most indoor environment, and can provide reference to the location department of related requirements.
【学位授予单位】：江西理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN925.93

【参考文献】