校园网络流量分流过程的分析与程序实现

随着近年来网络使用的普及,针对网络流量分析领域的研究越来越多.提出了各种流量分类算法,其最终目的是为了更好地识别网络带宽中的流量组成.基于对多种不同的流量分类算法的研究发现,网络流量的成流过程作为流量分类算法的基础性工作起着不可或缺的作用,并发现数据包成流的定义也各不相同.描述了一种基于数据包五元组的网络流量成流过程,对某校园网络流量实现了离线式的数据包成流,并针对成流过程中所出现的一些问题进行分析,给出可能的解决方案.

Implementation and Analysis of the Split-flow Process on the Campus Network Traffic

With the popularity of the network in recent years,researches aiming at the network traffic analysis increase sharply.All kinds of algorithms whose final goals are to identify the component of the network bandwidth are proposed.It is found that the process of split-flow considered as the fundamental work to the classification algorithms plays an irreplaceable role during the traffic classification based on the analysis on different traffic classification algorithms and with different definitions of the flow.Therefore,via an offline classification model of a certain campus network traffic,a process of the traffic split-flow based on the five-tuple of the packet and its several solutions are put forward to the problems during the split-flow process.