摘要
针对异构网络中拓扑环境复杂多变、负载不均衡等特性,传统的网络认知已经不能满足其性能发展的需求,所以合理的关键参数研究是异构网络性能提高的重要前提。在无对象模型和先验知识的情况下,使用ns2模拟大量具有瓶颈链路与不同状态的异构网络环境,提出两个新参数。使用格贴近度的方法深入分析各参数的相关性,新参数在分析中相关性较高,有效地反应出当前网络性能,也为分段连接的异构网络性能增强提供了可靠的基础。
Targeted at heterogeneous network, the environment of topology is complex, changeable, unbalanced load, etc. Traditional cognition of the network is no longer meeting the needs of its performance enhancing. So an important prereq-uisite for heterogeneous network performance enhancing is analyzing parameters reasonably. In the case of no object mod-el and prior knowledge, this paper uses the ns2 to simulate a large number of heterogeneous network environments with the bottleneck link and different state, then creates two new parameters. It uses the lattice close-degree method to analyse the correlation of each parameter depthly. The new parameters have high correlation and response the current network per-formance effectively, providing a reliable basis for future work on split-link connection of heterogeneous network perfor-mance enhancement.
出处
《计算机工程与应用》
CSCD
2014年第9期88-92,共5页
Computer Engineering and Applications
基金
国家自然科学基金(No.61070247)
河南省教育厅自然科学研究计划项目(No.2010A520017)
河南省科技攻关项目(No.112102210186)
关键词
异构网络
瓶颈链路
分段连接
格贴近度
相关性
heterogeneous network
bottleneck link
split-link connection
lattice close-degree
correlation
