The Implementation of Cluster-Based Traffic Simulation System

Self-similarity is one of the most important characteristics of real network traffic. Constructing an accurate model can reflect the real network environment precisely. It is also an effective measure to predict the upcoming network traffic, thus ensuring the quality and reliability of network service. In this paper, we present a cluster-based distributed system, which is based on FGN(Fraction Gaussian Noise) algorithm. With the purpose to generate large-scale traffic and simulate the real network environment. We offer a detailed description for this system in the following three aspects: algorithm, architecture and implementation. Experiments have proved that the cluster-based system can generate large-scale similar network traffic with designated Hurst parameter.