本文選題：數(shù)據(jù)挖掘技術(shù)　切入點：基站定位　出處：《湘潭大學(xué)》2017年碩士論文　論文類型：學(xué)位論文

【摘要】：科技的進步,生活水平的提高,數(shù)據(jù)時代的來臨,人們對于服務(wù)行業(yè)的需求也日益增加,這里就包括無線定位服務(wù)。因此,各種各樣的定位技術(shù)也得到了廣泛的發(fā)展,讓我們熟知的如:百度地圖,谷歌地圖,高德地圖等,都有一套非常成熟的定位技術(shù)。基站定位技術(shù)也在用戶移動終端定位方面有了長足的進步,成熟的定位技術(shù)也有它不可替代的重要意義和用途。本篇論文首先主要針對基站定位技術(shù)在國內(nèi)外發(fā)展的現(xiàn)狀,以及基站定位原理進行了概述,介紹了該定位技術(shù)與其他定位技術(shù)的問題與優(yōu)勢,為后面的優(yōu)化奠定了基礎(chǔ)。其次,深入剖析了指紋庫,分別介紹了指紋庫在前期離線數(shù)據(jù)采集階段和實時定位階段的原理,并分析了在使用指紋庫識別時候需要的指標變量,以及實現(xiàn)的一般過程。最后,簡單介紹了數(shù)據(jù)挖掘技術(shù)的一般過程。在移動終端定位之前,根據(jù)網(wǎng)絡(luò)優(yōu)化的業(yè)務(wù)需求對原始數(shù)據(jù)集,編寫算法進行了篩選。為了進一步提高定位的時效性,針對移動基站定位問題,提出了幾種優(yōu)化算法,分別闡述了這幾種算法的優(yōu)缺點和實現(xiàn)過程,最后做了結(jié)果對比分析,其中得到在使用BiK_means聚類算法時,對指紋庫數(shù)據(jù)集的聚類效果是比較理想的。因此在后面的定位當(dāng)中采取了該聚類算法,結(jié)合該聚類算法后,在資源利用率與用戶感知度分析時候,對業(yè)務(wù)進行分類,按照不同的業(yè)務(wù)指標分別對應(yīng)找出資源利用率與用戶感知度曲線的穩(wěn)定閾值。整體論文的思路是基于數(shù)據(jù)挖掘技術(shù)對指紋庫數(shù)據(jù)集進行聚類的過程。結(jié)合修正后的傳播模型,為了在以后的定位中達到更快捷,更省時的定位服務(wù),這也為以后得到實時的用戶數(shù)據(jù),進行用戶行為分析等方向的研究做了鋪墊工作。
[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.
【學(xué)位授予單位】：湘潭大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2017
【分類號】：TN929.5;TP311.13

【參考文獻】

相關(guān)期刊論文 前8條

1 王嘉e，

本文編號：1622605