论文部分内容阅读
准确刻画不同网络应用流量的行为特征,是识别和控制应用流量以及保证互联网服务质量的关键。该文提出一种基于流量系统状态递归特性的分析方法,研究应用流量的内在动力学行为。针对实际网络中多类不同应用流量的时间序列,首先重构流量序列的高维相空间,然后分析应用流量系统状态运动轨迹的递归特性,揭示其各自固有的内在行为。实验结果表明,流量的非线性动力学特征能够准确地刻画各类网络应用流量的行为,并且不随网络规模或网络协议版本的改变而发生变化。因此,流量的非线性动力学特征有助于提高互联网应用流量识别与控制的性能。
Accurately depict the behavior characteristics of different network application traffic is the key to identify and control application traffic and ensure Internet service quality. In this paper, an analysis method based on state recursion of traffic flow system is proposed to study the inherent dynamic behavior of traffic. Aiming at the time series of many kinds of different application traffic in the real network, the high dimensional phase space of the traffic sequence is reconstructed first, and then the recursive properties of the motion trajectory of the application traffic system are analyzed to reveal their intrinsic behaviors. The experimental results show that the nonlinear dynamic characteristics of traffic can accurately depict the traffic behavior of all kinds of networks and do not change with the change of network size or network protocol version. Therefore, the nonlinear dynamic characteristics of traffic help to improve the performance of Internet application traffic identification and control.