基于自回归滑动平均的网络数据流量预测模型

发布时间：2018-01-15 15:21

本文关键词：基于自回归滑动平均的网络数据流量预测模型　出处：《计算机科学》2014年04期 　论文类型：期刊论文

【摘要】：在无线网络中,对入侵攻击的准确和迅速的检测是关系到无线网络安全的重要问题。各种入侵攻击可以由其导致的网络流量的变化来检测。针对网络流量复杂的非线性以及混沌性,结合网络流量的时间序列特性,提出了一种基于自回归滑动平均(ARMA)的网络数据流量预测模型。该模型利用第三方检测系统,不需要耗费网络资源,能够迅速和准确地预测网络流量。采用从16个信道分析器获得的数据流量测量值对模型进行了初始化。仿真实验结果表明,文中提出的模型能够有效地检测网络入侵攻击,提高了整个网络的性能,延长了网络的寿命。
[Abstract]:On a wireless network. Accurate and rapid detection of intrusion attacks is an important issue related to wireless network security. All kinds of intrusion attacks can be detected by the changes of network traffic caused by intrusion attacks. Sex. Combined with the time series characteristics of network traffic, a network traffic prediction model based on autoregressive moving average (ARMA) is proposed, which utilizes a third-party detection system and does not need to consume network resources. The model is initialized using the data flow measurements obtained from 16 channel analyzers. The simulation results show that the proposed model can predict the network traffic quickly and accurately. The proposed model can detect intrusion attacks effectively, improve the performance of the whole network and prolong the lifetime of the network.
【作者单位】：沈阳大学信息中心;浙江大学计算机学院;
【基金】：国家航天局遥感论证中心项目(科工技2012A03A0939)资助
【分类号】：TN915.08
【正文快照】： 1引言WIA-PA无线网络标准[1-3]是一种工业自动化无线网络标准。它是为了满足工业自动化网络的特殊需求而制定的标准。这些特殊需求包括抗干扰、低能耗等。无线网络易于受到某些类型的网络攻击,因为其布局在公开和无保护的环境中。可以采用一些预先配置的安全系统提高无线网络

【参考文献】