论文部分内容阅读
在分析当前网络管理和主动网络技术的基础上,提出一种基于主动网络的分布式智能管理模型.采用多Agent 系统对其进行刻画与描述,设计用于主动网络中的 Agent 授权、管理和控制的管理控制报文(MCM).对于网络的故障诊断,采用证据理论实现不确定性推理,通过组合不同的故障征兆,以提高网络故障诊断的准确性.该模型具有分布式环境下的区域自治和多区域间协作等特点,对于大规模、异构网络的管理具有较明显的优势.
Based on the analysis of current network management and active network technology, a distributed intelligent management model based on active network is proposed, which is depicted and described by multi-agent system and designed for agent authorization, management and control in active networks (MCM) .For the network fault diagnosis, the theory of evidence is used to realize the uncertainty reasoning, which can improve the accuracy of network fault diagnosis by combining different fault symptoms.This model has the advantages of regional autonomy in distributed environment And multi-regional cooperation and other characteristics, for large-scale, heterogeneous network management has obvious advantages.