摘要
针对传统TCP控制协议在异构网络中效率低下问题,提出一种基于免疫和灰色系统理论的网络认知研究.该研究利用人工免疫系统分析各参数的特性和利用灰色关联度建立该网络各参数与当前网络性能之间的相关性,通过一定的数据分析,对网络进行准确、高效的认知.该研究采用传输控制协议TCPL进行仿真实验.仿真实验证明:该算法能够在不同背景流的条件下稳定发送窗口,减少拥塞发生并提高吞吐量,改善了网络性能.
In view of the traditional TCP control protocol in low efficiency in heterogeneous network, research of network cognitive based on immune and gray system theory is proposed. The research utilizes artificial immune system analysis parameters characteristics and the use of grey correlation to build the performance of the correlation between network parameters and the current network, through some data analysis to accurate and efficient cognition network. This research uses Transmission Control Protocol （TCPL） to finish simulation experiments. Simulation show： Under different background flow conditions, the algorithm can stabilize sending window, reduce congestion and improve throughput and transmission efficiency.
出处
《微电子学与计算机》
CSCD
北大核心
2013年第6期117-121,共5页
Microelectronics & Computer
基金
国家自然科学基金(61070247)
河南省教育厅自然科学研究计划项目(2010A520017)
河南省科技攻关项目(112102210186)
关键词
异构网络
人工免疫系统
灰色系统理论
瓶颈链路队列长度
网络认知
heterogeneous network
artificial immune system
gray system theory
the bottleneck link queue＇slength
network cognitive
