| As the scale and application of computer network is increasing day by day, the network has become a more and more important part in people's daily life and work. But as the complexity and heterogeneity of network are increasing, virus and destruction of human factors are also getting more and more through network, which interrupt the network's running order seriously. In this case, how to identificate the abnormal flow to guarantee the network performance and provide a good network environment for users become a focus. This paper starts from the abnormal traffic implement mechanism, analyzes the development and application of abnormal traffic identifying technology, studies the common characters, and accomplishes the research described as below:1. Studying the mechnism of several main abnormal traffic, and their features and problems in flow identification.2. Proposing several abnormal traffic identification methods based on transport layer features. Introducing the methods' theoretical basis and relative knowledge about NetFlow data format.3. Designing and implementing an abnormal traffic identification and recognition system. Introducing the entire architecture, realization mechanism and composing modules in detail.4. Experiments in Internet environment with proposed method by this paper and evaluation to the experiment result. |
