基于云计算的P2P识别系统（云墙）架构研究

发布时间：2018-12-13 22:10

【摘要】：P2P技术已经出现多年，但P2P技术从出现至今一直受到各种版权机构、网络服务提供商的打压，其原因就在于该技术能够让所有接入用户共享数据。而用户共享的数据很多都是没有授权不得传播的，而且各个P2P节点在传输和下载数据的时候会产生很多大流量，网络中大量的P2P流量会占用网络带宽导致网络拥塞，这无疑给网络服务运营商带来了麻烦。正是由于这些原因，P2P从出现至今一直被排挤打压。云计算是近几年新出现的一种技术，该技术从一出现就得到了广泛的关注，各个国家和世界知名IT公司也都投入了大量的人力物力对其进行研究和发展。由于云计算技术要依赖良好的网络才能运作，所以近几年为了促进云计算顺利发展各大公司和网络运营商都在积极改善网络环境。而网络也正是限制P2P发展的主要原因之一，所以云计算的发展在无形中促进了P2P的发展，近年来基于P2P技术的视频语音软件层出不穷，无论是P2P的文件共享还是P2P技术的实时通信都会产生巨大的网络流量，为了对P2P流量进行监控，P2P流量识别就成了亟待解决的问题。目前已经有了许多的P2P识别技术，但是面对着与日俱增的P2P流量，，无论是识别技术本身还是承载这些技术的硬件设备都开始显示出能力不足的问题。针对上述问题，本文研究了一种将单一设备上的P2P识别移植到云中的系统，该系统利用云的分步并行计算的特点将P2P流量识别分片细化，而且云计算本身的超强计算能力也要远远高于单一硬件设备，这使得现有的P2P识别技术能够轻松识别与日俱增的P2P流量成为可能。本文在第四章详细介绍了系统的组织架构，并用数学分析的方法推导出了本系统确实在识别效率方面有了明显的提升。
[Abstract]:P2P technology has been developed for many years, but since its emergence, P2P technology has been suppressed by various copyright agencies and network service providers. The reason is that the technology enables all access users to share data. Many of the data shared by users are not allowed to propagate without authorization, and many P2P nodes will generate a lot of traffic when they transmit and download data. A large amount of P2P traffic in the network will occupy the network bandwidth and lead to network congestion. This undoubtedly brings trouble to network service operator. It is for these reasons that P2P has been squeezed and suppressed since its emergence. Cloud computing is a new technology in recent years, which has been widely concerned since its emergence. IT companies in various countries and the world have also invested a lot of manpower and material resources to research and develop it. Cloud computing technology depends on good network to operate, so in recent years, in order to promote the smooth development of cloud computing, companies and network operators are actively improving the network environment. Network is one of the main reasons for restricting P2P development, so the development of cloud computing has promoted the development of P2P. In recent years, video and voice software based on P2P technology has emerged in endlessly. In order to monitor P2P traffic, P2P traffic identification becomes an urgent problem, both in P2P file sharing and P2P real-time communication. At present, there are many P2P identification technologies, but in the face of increasing P2P traffic, both the identification technology itself and the hardware equipment carrying these technologies are beginning to show the problem of insufficient capacity. In order to solve the above problems, this paper studies a system of transplanting P2P identification on a single device to cloud. The system makes use of the characteristics of two-step parallel computing of cloud to divide P2P traffic identification into pieces. Moreover, cloud computing itself has much higher computing power than a single hardware device, which makes it possible for the existing P2P identification technology to easily identify the increasing P2P traffic. In the fourth chapter, the organizational structure of the system is introduced in detail, and it is deduced by mathematical analysis that the recognition efficiency of the system has been improved obviously.
【学位授予单位】：江西理工大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP393.02

【参考文献】