基于用户社会性的定位方法的研究

发布时间：2018-01-20 04:50

本文关键词： 无线网络定位隐马尔可夫模型　出处：《上海交通大学》2013年硕士论文　论文类型：学位论文

【摘要】：随着无线网络的日益普及以及GPS[1]系统的广泛使用,无线网络中用户的位置信息越发重要。目前各种互联网产品都会推出基于LBG(Location Based Service)的服务,其中无线设备用户的物理位置信息是这些服务的基础。然而普遍使用着的GPS系统存在着很多问题,如在一些有遮蔽物的应用场景中无法使用,因而无线网络中的定位问题仍然有很大的空间供人们研究。目前被广泛研究的定位方法根据其所利用的测量信息可以被分为两种[16],一种是以细粒度[10]的精确测量值为主要依据的基于距离的定位方法；另一种是以粗粒度的较不精确的测量值为主要依据的基于连接信息[31]的定位方法。对于基于距离的定位方法,其主要思路是依靠获得的精确测量值计算无线信号发射源与接收源之间的距离,之后求解其相对位置。对于基于连接信息的定位方法,其主要思路是依靠获得的连接信息判断无线信号发射源与接收源距离间的关系,之后求解各节点可能的范围,最后根据相应目标函数得到其相对位置。我们通过对大规模无线网络用户数据的分析,发现用户在移动过程中具有移动范围稳定性、移动过程时空相关性以及社会相关性。并设计了基于这些特性的定位方法SOMA。最终通过基于真实的无线网络移动数据的实验验证了SOMA的性能。实验发现,SOMA的定位精确度比已知的定位方案的精确度高了不少,以平均绝对误差MAE为指标,MDS[30]的MAE值比SOMA高了近6倍；TSLRL[24]比SOMA高了近10倍,TSL比SOMA高了近10倍,LRL比SOMA高了近11倍,而Centroid[12]方法比SOMA高了近5倍。实验充分证明了我们的发现，即用户移动过程中存在着很强的社会相关性，对其加以利用可以极大的提高用户的定位的精确度。就我们目前的知识，，这是第一次有人将该特性应用于无线网络用户的定位问题中来。
[Abstract]:With the increasing popularity of wireless networks and GPS. [(1) widespread use of the system. The location information of users in wireless networks is becoming more and more important. At present, a variety of Internet products will launch services based on LBG(Location Based Service. The physical location information of wireless device users is the basis of these services. However, there are many problems in the commonly used GPS system, such as in some application scenarios with shelter. Therefore, there is still a lot of space for people to study the localization problem in wireless networks. Currently widely studied localization methods can be divided into two types according to the measurement information they use. [16], one is fine grained. [10] based on distance based positioning methods based mainly on precise measurements of coarse-grained values; another method based on connection information based on less precise measurements of coarse granularity. [The main idea of this method is to calculate the distance between the wireless signal transmitting source and the receiving source based on the obtained accurate measurement value. Then the relative position is solved. For the location method based on the connection information, the main idea is to judge the relationship between the radio signal transmitting source and the receiving source distance based on the connection information obtained. Then the possible range of each node is solved and its relative position is obtained according to the corresponding objective function. By analyzing the user data of large-scale wireless network, we find that the user has the stability of mobile range in the process of moving. Finally, a localization method based on these characteristics is designed. Finally, the performance of SOMA is verified by experiments based on real wireless network mobile data. The experimental results show that the accuracy of SOMA is much higher than that of the known positioning scheme, and the average absolute error (MAE) is taken as the index. [The MAE value of 30] is nearly 6 times higher than that of SOMA. [24] TSL is nearly 10 times higher than SOMA, nearly 10 times higher than SOMA, 11 times higher than SOMA, and Centroid. [The method is nearly 5 times higher than SOMA. The experiment fully proves our finding that there is a strong social correlation in the process of user movement. It is the first time that this feature has been applied to the location problem of wireless network users for the first time as far as our current knowledge is concerned.
【学位授予单位】：上海交通大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：TN929.5;P228.4

【相似文献】