基于QoS的电信网络海量数据处理与研究
发布时间：2019-05-28 02:40
【摘要】：移动互联网应用的飞速发展使得电信网络正面临着巨大的挑战，传统的服务质量（QoS）分析方法和工具难以解决这些问题。本论文针对基于QoS的电信网络海量数据分析和处理进行了研究，着重对于影响用户体验的电信网络业务进行了分析，并基于海量数据进行统计，分析给出了影响网络性能和用户体验的主要因素。 论文首先概述了服务质量的概念，并对服务质量的模型和关注的问题进行了讨论，分析给出了服务质量与体验质量的区别与联系。论文研究了基于数据特征的业务分类方法，，对移动互联网服务质量的关键因素进行了分析，并对影响服务质量的关键问题进行了深入研究，提出了解决方案。论文基于全网端到端的数据流量，首先从传统资源消耗的角度对全网进行了分析；又通过关键KPI指标，对不同类型的应用进行数据流特性统计分析，得到2G、3G下数据流特征规律。论文最后提出了一种基于Kmeans算法的拥塞避免机制，可以通过分析占据基站的流量特性，指导网络优化和提高网络服务质量。
[Abstract]:With the rapid development of mobile Internet applications, telecom networks are facing great challenges. Traditional quality of service (QoS) analysis methods and tools are difficult to solve these problems. In this paper, the analysis and processing of massive data in telecom network based on QoS are studied, with emphasis on the analysis of telecom network services that affect the user experience, and the statistics are carried out based on massive data. The main factors affecting network performance and user experience are analyzed. Firstly, this paper summarizes the concept of quality of service, discusses the model of quality of service and the problems concerned, and analyzes the difference and relationship between quality of service and quality of experience. In this paper, the service classification method based on data characteristics is studied, the key factors of mobile Internet quality of service are analyzed, the key problems that affect the quality of service are deeply studied, and the solutions are put forward. Based on the end-to-end data flow of the whole network, this paper first analyzes the whole network from the point of view of traditional resource consumption. Through the key KPI index, the data flow characteristics of different types of applications are statistically analyzed, and the characteristics of 2G, 3G data flow are obtained. Finally, a congestion avoidance mechanism based on Kmeans algorithm is proposed, which can guide the network optimization and improve the quality of network service by analyzing the traffic characteristics of the occupied base station.
【学位授予单位】：南京邮电大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP393.09
本文编号：2486665
[Abstract]:With the rapid development of mobile Internet applications, telecom networks are facing great challenges. Traditional quality of service (QoS) analysis methods and tools are difficult to solve these problems. In this paper, the analysis and processing of massive data in telecom network based on QoS are studied, with emphasis on the analysis of telecom network services that affect the user experience, and the statistics are carried out based on massive data. The main factors affecting network performance and user experience are analyzed. Firstly, this paper summarizes the concept of quality of service, discusses the model of quality of service and the problems concerned, and analyzes the difference and relationship between quality of service and quality of experience. In this paper, the service classification method based on data characteristics is studied, the key factors of mobile Internet quality of service are analyzed, the key problems that affect the quality of service are deeply studied, and the solutions are put forward. Based on the end-to-end data flow of the whole network, this paper first analyzes the whole network from the point of view of traditional resource consumption. Through the key KPI index, the data flow characteristics of different types of applications are statistically analyzed, and the characteristics of 2G, 3G data flow are obtained. Finally, a congestion avoidance mechanism based on Kmeans algorithm is proposed, which can guide the network optimization and improve the quality of network service by analyzing the traffic characteristics of the occupied base station.
【学位授予单位】：南京邮电大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP393.09
【参考文献】
相关期刊论文 前4条
1 林闯,单志广,盛立杰,吴建平;Internet区分服务及其几个热点问题的研究[J];计算机学报;2000年04期
2 曹志宇;张忠林;李元韬;;快速查找初始聚类中心的K_means算法[J];兰州交通大学学报;2009年06期
3 郑侃;张月莹;王文博;;浅析无线网络的用户体验质量建模及性能优化[J];信息通信技术;2012年03期
4 张健沛;杨悦;杨静;张泽宝;;基于最优划分的K-Means初始聚类中心选取算法[J];系统仿真学报;2009年09期
本文编号：2486665
本文链接：https://www.wllwen.com/guanlilunwen/ydhl/2486665.html
