本文選題：蜂窩無線網(wǎng)絡(luò) + 聚類分析��； 參考：《北京郵電大學(xué)》2016年碩士論文

【摘要】：隨著LTE技術(shù)的高速發(fā)展與廣泛應(yīng)用,移動用戶的行為規(guī)律也隨之發(fā)生了很大變化。無線技術(shù)的飛速發(fā)展,帶來的是移動終端使用頻率增大,使用時長變多。而無線數(shù)據(jù)的爆發(fā)增長,帶來的是日益增多的運(yùn)營機(jī)遇和數(shù)據(jù)業(yè)務(wù),與此同時,移動端數(shù)據(jù)的增長趨勢使得運(yùn)營商原本對網(wǎng)絡(luò)資源優(yōu)化難度不斷增大,因而采用新的分析方法,即同時分析用戶、業(yè)務(wù)、資源多維度的數(shù)據(jù)對于無線網(wǎng)絡(luò)規(guī)劃和資源優(yōu)化有重大意義。LTE網(wǎng)絡(luò)承載移動互聯(lián)網(wǎng)數(shù)據(jù)業(yè)務(wù)的逐步發(fā)展,為了應(yīng)對數(shù)據(jù)種類和數(shù)量激增的海量通信數(shù)據(jù),本文將數(shù)據(jù)挖掘的分析方法應(yīng)用到蜂窩無線網(wǎng)絡(luò)數(shù)據(jù)處理與分析中。另一方面,用戶通過移動設(shè)備接入互聯(lián)網(wǎng)的頻率日益增長,搭載安卓操作系統(tǒng)的設(shè)備的占有率超過80%,安卓系統(tǒng)具有極好的代表性�；谝陨媳尘�,本論文首先對蜂窩無線網(wǎng)絡(luò)中的用戶、業(yè)務(wù)、資源數(shù)據(jù)研究分析,并給出一種基于聚類分析的多維度擬合關(guān)聯(lián)預(yù)測方法。在此之后搭建了一個基于安卓的移動互聯(lián)網(wǎng)終端分析系統(tǒng),實(shí)現(xiàn)了從用戶端采集數(shù)據(jù),并使用聚類算法分析,將結(jié)果展示在web端的整體流程。本論文主要內(nèi)容如下:1.給出了一個基于聚類分析的多維度擬合關(guān)聯(lián)預(yù)測方法。在該方法中,首先通過主成分分析,將待分析樣本數(shù)據(jù)進(jìn)行降維處理,隨后通過二次聚類的方法對數(shù)據(jù)聚類分析,并將結(jié)果采用梯度擬合,進(jìn)行蜂窩無線網(wǎng)絡(luò)數(shù)據(jù)的關(guān)聯(lián)預(yù)測。研究結(jié)果表明:與對數(shù)據(jù)直接擬合相比,聚類分析能夠提高對大量LTE數(shù)據(jù)預(yù)測的精度,實(shí)現(xiàn)資源預(yù)測方面的優(yōu)化。2.設(shè)計(jì)了基于安卓的移動互聯(lián)網(wǎng)終端分析系統(tǒng)。該系統(tǒng)包括基于安卓終端的數(shù)據(jù)采集系統(tǒng)和服務(wù)器端的數(shù)據(jù)分析系統(tǒng)兩部分。在安卓終端的采集系統(tǒng)可以采集蜂窩用戶的多種數(shù)據(jù),包括用戶基本信息、信號數(shù)據(jù)、通話數(shù)據(jù)、短信數(shù)據(jù)、流量業(yè)務(wù)數(shù)據(jù)以及通信異常數(shù)據(jù)等,同時終端記錄發(fā)生業(yè)務(wù)時的GPS坐標(biāo)。服務(wù)器端的分析系統(tǒng)存儲終端采集的數(shù)據(jù),實(shí)現(xiàn)本文的聚類算法,并使用現(xiàn)網(wǎng)真實(shí)數(shù)據(jù)進(jìn)行分析預(yù)測。本文設(shè)計(jì)的分析系統(tǒng),結(jié)合本文給出的方法,為蜂窩無線網(wǎng)絡(luò)的數(shù)據(jù)采集、分析,預(yù)測提供了一個完整的解決流程。
[Abstract]:With the rapid development and wide application of LTE technology, the behavior of mobile users has changed greatly. With the rapid development of wireless technology, the frequency of mobile terminal is increasing and the time is long. The explosion and growth of wireless data brings more and more operation opportunities and data services. At the same time, the increasing trend of mobile data makes it more difficult for operators to optimize network resources, so a new analysis method is adopted. That is to say, it is of great significance for wireless network planning and resource optimization to analyze the multi-dimensional data of users, services and resources at the same time. LTE network carries the gradual development of mobile Internet data services. In order to deal with the huge amount of communication data, the analysis method of data mining is applied to the data processing and analysis of cellular wireless network. On the other hand, the frequency of users accessing the Internet through mobile devices is increasing, with more than 80% of devices running Android operating system, which is an excellent representative. Based on the above background, this paper first studies and analyzes the data of users, services and resources in cellular wireless networks, and presents a multi-dimensional fitting association prediction method based on clustering analysis. After that, a mobile Internet terminal analysis system based on Android is built, which can collect data from the user side and use clustering algorithm to analyze the result. The result is displayed in the whole flow of web. The main contents of this thesis are as follows: 1. A multi-dimensional fitting association prediction method based on cluster analysis is presented. In this method, firstly, the sample data are reduced by principal component analysis, then the data are clustered by quadratic clustering method, and the results are fitted by gradient to predict the data association of cellular wireless network. The results show that the clustering analysis can improve the precision of forecasting a large number of LTE data and realize the optimization of resource prediction by comparing with the direct fitting of the data. A mobile internet terminal analysis system based on Android is designed. The system includes two parts: the data acquisition system based on Android terminal and the data analysis system on the server side. The collection system of Android terminal can collect a variety of data from cellular users, including user basic information, signal data, call data, short message data, traffic data and abnormal communication data, etc. At the same time, the terminal records the GPS coordinates when the business occurs. The server side analysis system stores the data collected by the terminal, realizes the clustering algorithm of this paper, and uses the real data of the existing network to analyze and predict. The analysis system designed in this paper, combined with the method given in this paper, provides a complete solution for the data acquisition, analysis and prediction of cellular wireless network.
【學(xué)位授予單位】：北京郵電大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2016
【分類號】：TN929.5;TP311.52

【參考文獻(xiàn)】

相關(guān)期刊論文 前7條

1 繆其浩;;大數(shù)據(jù)時代:趨勢和對策[J];科學(xué);2013年04期

2 王克曉;李鳳友;劉煥玲;邢衛(wèi)軍;任洪文;;手持GPS定位精度與誤差的研究[J];全球定位系統(tǒng);2011年06期

3 陳文周;;WiFi技術(shù)研究及應(yīng)用[J];數(shù)據(jù)通信;2008年02期

4 邵長彬;李洪亮;;用Busybox制作嵌入式Linux根文件系統(tǒng)[J];微計(jì)算機(jī)信息;2007年29期

5 彭文波;;利用TCPDump檢測網(wǎng)絡(luò)數(shù)據(jù)[J];計(jì)算機(jī)安全;2006年07期

6 姚曉宇,趙晨;Linux內(nèi)核防火墻Netfilter實(shí)現(xiàn)與應(yīng)用研究[J];計(jì)算機(jī)工程;2003年08期

7 鄧平,李莉,范平志;一種TDOA/AOA混合定位算法及其性能分析[J];電波科學(xué)學(xué)報;2002年06期

相關(guān)博士學(xué)位論文 前1條

1 劉鵬;網(wǎng)絡(luò)用戶行為分析的若干問題研究[D];北京郵電大學(xué);2010年

，

本文編號：1841170