【摘要】：21世紀是信息的時代,信息交換已經成為人類生活最重要的組成之一。網絡作為信息交換的主要載體,已經成為社會最重要的基礎設施。移動互聯(lián)網、物聯(lián)網和車聯(lián)網等概念的興起,證明網絡已經成為社會生活的核心；教育學術、商業(yè)經濟和娛樂文化等方方面面都離不開網絡的支撐。隨著人們對網絡越來越依賴,網絡的規(guī)模也越來越龐大,人們對網絡性能的要求也越來越高。如何保證規(guī)模龐大的網絡的有效性、可靠性和穩(wěn)定性,如何科學高效的管理網絡,以及如何在當前網絡技術上演進出更加先進的網絡技術已經成為眾多學者和研究機構的研究熱點。而這些問題的解決依賴對網絡進行有效的測量,并通過計算測量數(shù)據來分析網絡性能。 本文將論述云監(jiān)測平臺的海量測量數(shù)據分析技術,旨在通過分析大規(guī)模網絡測量得到的海量測量數(shù)據來評價當前被測量網絡性能。本文論述的測量數(shù)據分析技術分為測量數(shù)據分析算法和測量數(shù)據分析系統(tǒng)兩部分。測量數(shù)據分析算法提供一種海量測量數(shù)據處理方法,通過對海量的測量數(shù)據進行運算來分析被測量網絡的性能狀態(tài)。測量數(shù)據分析系統(tǒng)是對測量數(shù)據分析算法的系統(tǒng)實現(xiàn)。測量數(shù)據分析系統(tǒng)采用云計算編程模型對分析算法進行實現(xiàn),旨在依托云計算強大的存儲和計算能力,向網絡管理人員提供一種快速處理海量測量數(shù)據的工具。 本文首先介紹云監(jiān)測平臺的海量測量數(shù)據分析技術的研究背景和研究意義,并對當前的研究技術進行介紹,其中包括了現(xiàn)有的網絡測量方法和性能評價方法。其次,本文將討論測量數(shù)據分析系統(tǒng)的設計需求和系統(tǒng)需求,確定測量數(shù)據分析系統(tǒng)的設計目標、實現(xiàn)原則、功能需求和性能需求。然后,本文將著重介紹測量數(shù)據分析算法,其中包括通過計算海量測量數(shù)據分析路徑性能的綜合指標法和能夠準確分析網絡性能的低耦合網絡性能分析法。之后,本文將討論測量數(shù)據分析系統(tǒng)整體結構和各部分結構,從系統(tǒng)實現(xiàn)角度介紹測量數(shù)據分析系統(tǒng)。最后,本文采用幾組實驗來介紹云監(jiān)測平臺的海量測量數(shù)據分析技術在實際場景中的應用,以實例證明其在實際應用中的價值。
[Abstract]:The 21st century is the era of information, information exchange has become one of the most important components of human life. As the main carrier of information exchange, network has become the most important infrastructure of society. The rise of the concepts of mobile Internet, Internet of things and car networking proves that the network has become the core of social life; the education, academic, business economy and entertainment culture can not be separated from the support of the network. As people rely more and more on the network, the scale of the network is becoming larger and larger, and the demand for network performance is becoming higher and higher. How to ensure the effectiveness, reliability and stability of the large-scale network, how to manage the network scientifically and efficiently, And how to evolve more advanced network technology in the current network technology has become the research hotspot of many scholars and research institutions. The solution of these problems depends on the effective measurement of the network and the analysis of the network performance by calculating the measurement data. This paper will discuss the massive measurement data analysis technology of cloud monitoring platform, which aims to evaluate the performance of the measured network by analyzing the massive measurement data obtained from the large-scale network measurement. The technique of measuring data analysis is divided into two parts: measurement data analysis algorithm and measurement data analysis system. The measurement data analysis algorithm provides a massive measurement data processing method to analyze the performance state of the measured network through the operation of the massive measurement data. The measurement data analysis system is the system realization of the measurement data analysis algorithm. The measurement data analysis system uses the cloud computing programming model to implement the analysis algorithm, which aims to provide network managers with a tool to deal with massive measurement data quickly, relying on the powerful storage and computing capabilities of cloud computing. This paper first introduces the research background and significance of cloud monitoring platform mass measurement data analysis technology, and introduces the current research technology, including the existing network measurement methods and performance evaluation methods. Secondly, this paper will discuss the design requirements and system requirements of the measurement data analysis system, and determine the design objectives, realization principles, function requirements and performance requirements of the measurement data analysis system. Then, this paper will focus on the measurement data analysis algorithm, including the comprehensive index method to analyze the path performance of mass measurement data and the low-coupling network performance analysis method which can accurately analyze the network performance. After that, this paper will discuss the whole structure and each part structure of the measurement data analysis system, and introduce the measurement data analysis system from the point of view of the system implementation. Finally, this paper uses several experiments to introduce the application of cloud monitoring platform's massive measurement data analysis technology in the actual scene, and proves its value in practical application by an example.
【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】：TP393.0

【參考文獻】

相關期刊論文 前10條

1 畢經平,李忠誠,吳起;大規(guī)�；ヂ�(lián)網端到端行為評價指標研究[J];計算機工程與應用;2002年12期

2 談杰;李星;;網絡測量綜述[J];計算機應用研究;2006年02期

3 張宏莉,方濱興,胡銘曾,姜譽,詹春艷,張樹峰;Internet測量與分析綜述[J];軟件學報;2003年01期

4 蔡志平;劉芳;趙文濤;劉湘輝;殷建平;;網絡測量部署模型及其優(yōu)化算法[J];軟件學報;2008年02期

5 楊雅輝,李小東;IP網絡性能指標體系的研究[J];通信學報;2002年11期

6 紀其進,董育寧;IP網絡性能特征模型分析[J];通信學報;2004年03期

7 張冬艷;胡銘曾;張宏莉;;基于測量的網絡性能評價方法研究[J];通信學報;2006年10期

8 趙娟;郭平;鄧宏鐘;吳俊;譚躍進;李建平;;基于信息流動力學的通信網絡性能可靠性建模與分析[J];通信學報;2011年08期

9 張秀武;雷為民;;基于網絡性能的VoIP語音質量評價模型[J];小型微型計算機系統(tǒng);2010年03期

10 何娜;崔毅東;楊談;金躍輝;;基于云計算的分布式網絡測量系統(tǒng)設計[J];微計算機應用;2011年11期

，

本文編號：2419344