一个网络管理软件中基于自适应的流量预测  

Traffic Forecast Based on Self-adaptive in a Network Management System

作　　者：姚奇富[1] 马华林[1] 张国清[2] 

机构地区：[1]浙江工商职业技术学院,浙江宁波315012 [2]中国科学院计算技术研究所,北京100080

出　　处：《计算机测量与控制》2006年第1期45-46,66,共3页Computer Measurement &Control

基　　金：国家863基金资助项目(2002AA142030)

摘　　要：传统的基于流量的网络报警系统往往根据系统配置的固定阈值来判断当前网络流量是否正常,通过运用统计学和预测学对网络流量预测方法进行研究,提出了一种能根据网络实际流量状态自己调整阈值的新型网络报警机制,能有效地发现网络流量异常,从而为评估和优化网络性能提供有力的依据。Traditional network alarm system which based on traffic usually sets fixed threshold to determine whether the current network traffic are at normal situation. After researching on network traffic forecasts by using statistical arithmetic method, a new alarm mechanism is put forward, which will adjust the threshold automatically as well as send alarm messages according to the real network traffic states. Such kind of mechanism will find the abnormal states of the network traffic, sequentially provide strong evidences to evaluate and optimize it.

关 键 词：网络管理 网络性能 流量预测 报警 

分 类 号：TP316.8[自动化与计算机技术—计算机软件与理论]