基于DPI技术的VoIP流量识别与控制

发布时间：2018-06-19 21:43

本文选题：流量检测 + 流量控制　；参考：《江西理工大学》2014年硕士论文

【摘要】：VoIP(Voice over Internet Protocol)是一种新兴起的互联网通讯技术。因为VoIP具有成本低和开放灵活的特点，它对传统的电信运营商产生了巨大的威胁，尤其对于国内外长途通话业务，已经产生了很多的利益损失。由于VoIP能够带来很大的利润以及国家监管放松的原因，很多非法的运营商开始经营VoIP业务，造成传统电信运营商的利润减少；国家对于PC To Phone业务的监管还处于未开始状态，当有突发状况发生时，国家无法保证通信系统的安全；很多虚拟运营商已经在国内开展业务，但是，国家对于他们还没有开始进行监管，也没有相关的管理经验。所以，国家需要研究相关的技术用来应对VoIP对于传统电信运营商的威胁。我们需要准确识别出网络中的VoIP流量才能进行管理。本文重点研究了VoIP网络流量检测的关键技术，深入分析了多种网络流量识别技术，对AC算法进行了分析研究，并且在这个基础上设计实现了一个网络流量监控系统，可以达到对VoIP网络流量的检测和控制。本文的主要内容包括下面几个部分： (1)研究分析了网络流量识别的关键技术——DPI(深度包检测技术)和DFI(深度/动态流检测技术)，比较了这两种流量检测技术的优缺点，并合理利用两种技术的结合来实现网络流量的识别。然后也研究了网络流量的控制技术，对网络流量控制技术进行了比较分析。 (2)对DPI技术的理论基础——模式匹配算法进行了研究，重点研究了AC自动机算法，，分析了AC自动机在DPI检测中的应用。 (3)在研究分析网络流量识别和控制的关键技术之后，提出了一种基于DPI的网络流量识别和控制技术解决方案，设计实现了一种网络流量识别和控制系统，详细分析了流量识别模块的实现、识别引擎的工作原理以及数据包的检测方式，并介绍了DPI特征库，最后对系统的网络部署方案做了详细的介绍，提出了一种改进的网络部署方案。 (4)该网络流量监控系统需要定期升级特征库，用来应对网络技术的快速发展。经过测试，该系统网络流量识别率高，能够识别出一百多种VoIP网络应用软件，对VoIP流量的阻断测试也很成功，该系统在稳定性和可用性上表现良好。
[Abstract]:VoIP Voice over Internet Protocol is a new Internet communication technology. Because VoIP has the characteristics of low cost and open and flexible, it poses a great threat to traditional telecom operators, especially for long-distance call services at home and abroad, which has caused a lot of loss of interest. Because VoIP can bring great profits and the state regulation is relaxed, many illegal operators begin to operate VoIP business, resulting in a decrease in the profits of traditional telecom operators, and the state regulation of PC to phone business is still in a state of not beginning. When an emergency occurs, the state cannot guarantee the security of the communication system; many virtual operators already operate in China, but the state has not started to regulate them, nor has the relevant management experience. Therefore, the state needs to study the relevant technologies to deal with VoIP threats to traditional telecom operators. We need to accurately identify VoIP traffic in the network to manage. This paper focuses on the key technologies of VoIP network traffic detection, deeply analyzes various network traffic identification technologies, analyzes the AC algorithm, and designs and implements a network traffic monitoring system on this basis. Can achieve VoIP network traffic detection and control. The main contents of this paper include the following parts: 1) the key technologies of network traffic identification, DPI (depth packet detection) and DFI (depth / dynamic flow detection), are analyzed and compared. The advantages and disadvantages of the detection technology, And reasonable use of the combination of the two technologies to achieve network traffic identification. Then, the control technology of network traffic is studied, and the network traffic control technology is compared and analyzed. (2) the pattern matching algorithm, which is the theoretical foundation of DPI technology, is studied, and the AC automata algorithm is studied emphatically. This paper analyzes the application of AC automata in DPI detection. After studying and analyzing the key technologies of network traffic identification and control, a solution of network traffic identification and control technology based on DPI is proposed. A network traffic identification and control system is designed and implemented. The realization of the flow identification module, the working principle of the identification engine and the detection method of the data packet are analyzed in detail, and the DPI signature library is introduced. Finally, the network deployment scheme of the system is introduced in detail, and an improved network deployment scheme is proposed. 4) the network traffic monitoring system needs to update the signature library periodically to cope with the rapid development of network technology. The network traffic identification rate of the system is high and more than 100 kinds of VoIP network applications can be identified. The VoIP traffic blocking test is also successful. The system performs well in stability and availability.
【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】：TP393.06

【参考文献】