【摘要】：近年來移動互聯(lián)網的迅速發(fā)展在為網民提供便捷的同時,已經深入滲透到人們的生活中。2013年1月15日中國互聯(lián)網絡信息中心(CNNIC)發(fā)布的《中國互聯(lián)網絡發(fā)展狀況統(tǒng)計報告》中顯示,移動互聯(lián)網各項指標增長速度全面超越傳統(tǒng)寬帶網絡。自我國開啟3G網絡服務以來,智能終端快速普及。2012年我國已經成為全球智能手機出貨量最大的國家。終端是用戶產生移動互聯(lián)網絡流量的主要載體,終端的發(fā)展引導用戶上網行為的變化,識別移動互聯(lián)網中的終端,掌握移動互聯(lián)網中不同終端用戶的上網行為、流量特征,對把握移動互聯(lián)網的發(fā)展具有重要意義。 本文首先介紹了移動互聯(lián)網的發(fā)展現(xiàn)狀,包括移動互聯(lián)網網絡流量分析的內容與終端識別方法。隨后還介紹了云計算與海量數據挖掘技術與Hadoop分布式處理框架,詳細介紹了實驗室內基于Hadoop的分布式大數據分析系統(tǒng)。 通過在實驗室的Hadoop大數據分析系統(tǒng)上對海量移動互聯(lián)網數據進行分析與研究,我們研發(fā)了挖掘海量上網數據中的終端型號以及操作系統(tǒng)算法。本文詳細介紹了基于Hadoop的終端型號與操作系統(tǒng)識別算法實現(xiàn)流程。 在研發(fā)終端型號識別與操作系統(tǒng)識別的基礎上,我們使用算法識別出的手機型號與操作系統(tǒng)圍繞終端對移動互聯(lián)網流量特征進行分析。最后,為了更明確的體現(xiàn)移動互聯(lián)網網絡流量特征,我們將圍繞終端分析的移動互聯(lián)網網絡流量特征與ADSL網絡流量特征進行對比分析。在介紹ADSL網絡流量特征的同時分析了ADSL網絡中五種典型業(yè)務的流量對網絡總流量的影響因子,并給出了流量預測方程。
[Abstract]:In recent years, the rapid development of the mobile Internet has provided convenience for Internet users at the same time, It has penetrated deeply into people's lives. According to the "Statistics report on the Development of China's Internet Network" released by the China Internet Information Center (CNNIC) on January 15, 2013, All indicators of mobile Internet growth rate surpasses the traditional broadband network. Since the start of 3G network service in China, smart terminals have been rapidly popularized. In 2012, China has become the largest country in terms of smartphone shipments in the world. Terminal is the main carrier of mobile Internet traffic generated by users. The development of terminals leads users to change their behavior on the Internet, identify the terminals in the mobile Internet, and master the behavior and traffic characteristics of different end-users in the mobile Internet. It is of great significance to grasp the development of mobile Internet. This paper first introduces the development of mobile Internet, including the content of mobile Internet network traffic analysis and terminal identification method. Then it also introduces cloud computing, mass data mining technology and Hadoop distributed processing framework, and introduces the distributed big data analysis system based on Hadoop in the laboratory in detail. By analyzing and researching the massive mobile Internet data on the Hadoop big data analysis system in the laboratory, we have developed the terminal model and the operating system algorithm of mining the massive Internet data. This paper introduces the realization flow of terminal model and operating system recognition algorithm based on Hadoop in detail. Based on the research and development of terminal model recognition and operating system recognition, we use the algorithm to identify the mobile phone model and operating system around the terminal to analyze the mobile Internet traffic characteristics. Finally, in order to reflect the characteristics of mobile Internet network traffic more clearly, we will compare and analyze the characteristics of mobile Internet network traffic and ADSL network traffic characteristics analyzed by terminals. In this paper, the traffic characteristics of ADSL network are introduced, and the influence factors of five typical traffic in ADSL network on the total network traffic are analyzed, and the traffic prediction equation is given.
【學位授予單位】：北京郵電大學
【學位級別】：碩士
【學位授予年份】：2014
【分類號】：TN929.53;TP393.06

【參考文獻】

相關期刊論文 前2條

1 何莘;王琬蕪;;自然語言檢索中的中文分詞技術研究進展及應用[J];情報科學;2008年05期

2 胡海波;王林;;冪律分布研究簡史[J];物理;2005年12期

，

本文編號：2366591