数据挖掘技术在无线定位中的应用

发布时间：2018-03-17 01:37

本文选题：数据挖掘技术　切入点：基站定位　出处：《湘潭大学》2017年硕士论文　论文类型：学位论文

【摘要】：科技的进步,生活水平的提高,数据时代的来临,人们对于服务行业的需求也日益增加,这里就包括无线定位服务。因此,各种各样的定位技术也得到了广泛的发展,让我们熟知的如:百度地图,谷歌地图,高德地图等,都有一套非常成熟的定位技术。基站定位技术也在用户移动终端定位方面有了长足的进步,成熟的定位技术也有它不可替代的重要意义和用途。本篇论文首先主要针对基站定位技术在国内外发展的现状,以及基站定位原理进行了概述,介绍了该定位技术与其他定位技术的问题与优势,为后面的优化奠定了基础。其次,深入剖析了指纹库,分别介绍了指纹库在前期离线数据采集阶段和实时定位阶段的原理,并分析了在使用指纹库识别时候需要的指标变量,以及实现的一般过程。最后,简单介绍了数据挖掘技术的一般过程。在移动终端定位之前,根据网络优化的业务需求对原始数据集,编写算法进行了筛选。为了进一步提高定位的时效性,针对移动基站定位问题,提出了几种优化算法,分别阐述了这几种算法的优缺点和实现过程,最后做了结果对比分析,其中得到在使用BiK_means聚类算法时,对指纹库数据集的聚类效果是比较理想的。因此在后面的定位当中采取了该聚类算法,结合该聚类算法后,在资源利用率与用户感知度分析时候,对业务进行分类,按照不同的业务指标分别对应找出资源利用率与用户感知度曲线的稳定阈值。整体论文的思路是基于数据挖掘技术对指纹库数据集进行聚类的过程。结合修正后的传播模型,为了在以后的定位中达到更快捷,更省时的定位服务,这也为以后得到实时的用户数据,进行用户行为分析等方向的研究做了铺垫工作。
[Abstract]:With the development of science and technology, the improvement of living standard, the arrival of data age, the demand for service industry is increasing day by day, including wireless location service. Let us be familiar with such as: Baidu Maps, Google Maps, Amap and so on, all have a set of very mature positioning technology. Base station positioning technology has also made great progress in user mobile terminal positioning. Mature positioning technology also has its irreplaceable significance and use. Firstly, this paper mainly focuses on the development of base station positioning technology at home and abroad, as well as the basic principles of base station positioning. This paper introduces the problems and advantages of the location technology and other positioning techniques, which lays a foundation for the later optimization. Secondly, the fingerprint database is deeply analyzed, and the principles of the fingerprint database in the early off-line data acquisition stage and the real-time positioning stage are introduced, respectively. At last, the general process of data mining technology is introduced briefly. In order to further improve the timeliness of location, several optimization algorithms are proposed to solve the problem of mobile base station location. The advantages and disadvantages of these algorithms and the implementation process are described respectively. Finally, the results are compared and analyzed, and the results are obtained when using BiK_means clustering algorithm. The clustering effect of fingerprint database data set is ideal. Therefore, the clustering algorithm is adopted in the later location. After combining the clustering algorithm, the business is classified when the resource utilization ratio and the user perception are analyzed. According to different business indexes, the stable threshold of resource utilization ratio and user perception curve is found respectively. The whole idea of this paper is the process of clustering fingerprint database data set based on data mining technology, combined with the modified propagation model. In order to achieve a faster and more time-saving positioning service in the future, it also paves the way for the research of real-time user data and user behavior analysis in the future.
【学位授予单位】：湘潭大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN929.5;TP311.13

【参考文献】