摘要
网络流量模型以考察网络流量特性为出发点,以数学理论为基础,通过建立数学模型来反映真实的网络流量及其变化趋势。传统的泊松模型在现代数据网络中已经不再适用,不能真实地反映流量的趋势。但是自从网络流量的自相似性被发现后,网络流量的自相似模型不断涌现。文中应用了既能反映长相关性又能反映短相关性的FARIMA模型对真实网络流量数据进行了分析预测,经过研究和实践的验证,对模型进行了改进,提出了SFARIMA网络流量预测模型。
By collecting and analyzing the characteristics of network traffic,network traffic models can reflect the real network traffic using mathematical methods.The traditional Poisson model is not suitable in the modern data networks which can not reflect the real flow trend.However,since Hosking discovers the self-similarity of network traffic,self-similar models have continued to emerge. This article focus on researching network traffic data analysis and forecasting using FARIMA model.The experiment data is quite similar to the real one,because it's not only LRD but also SRD.And the model has been improved,this article puts forward a new network traffic forecasting model called SFARIMA.
出处
《计算机技术与发展》
2010年第12期54-56,188,共4页
Computer Technology and Development
基金
国家高技术研究发展计划(863)资助项目(2009AA01Z202
2009AA01Z212)