摘要
根据拓展仿真的原理和方法,以建立符合网络特性的流量预测模型为目标,笔者提出一种基于定量统计建模方法的网络仿真平台.结合网络仿真器NS构建基于DWT和FARIMA的流量预测模型,并应用该模型的预测结果辅助主动队列管理的显式拥塞标记,采取拥塞预警的方式达到对网络协议进行优化的目的.文中给出所设计的分层组播协议FLM实现算法,并开展了关于流量行为和协议策略的仿真测试.
In analyzing model is order to monitor and predict the behavior properties of network traffic flow, a quantiative proposed,in which the DWT and FARIMA scheme based on NS2 simulator is adopted to assist the AQM's ECN marking strategies. The network protocol can thus be optimized with the indication of congestion caution information. Detailed algorithm of FLM designed according to the model is presented, and simulating experiments are also performed to test and demonstrate the effects of traffic analysis and protocol optimization.
出处
《山东师范大学学报(自然科学版)》
CAS
2011年第3期36-40,共5页
Journal of Shandong Normal University(Natural Science)
基金
国家科技重大专项(2009ZX07210-007)
山东省高等学校科技计划项目:网络流量建模与预测技术研究(J09LG75)
山东师范大学实验教学改革立项:定量分析模拟实验平台的构建.
关键词
仿真
定量分析建模
网络拥塞
协议优化
simulation
quantitative analysis modeling
network congestion
protocol optimization