摘要
本文利用神经网络方法和粗集理论方法,从原始网络管理信息库MIB中提取网络故障征兆信息,自动创造和维护网络故障诊断知识库,实现网络故障的提前告警和故障原因的关联分析。利用这一方法可以从网络管理信息库中学习和了解网络系统的深层次行为。
Modern computer networks increasing in size and complexity require intelligent techniques for managing and maintain them. Based on ANN technique and Rough Theory, we present an approach which can obtain and analyze the network management information and map the raw network fault symptoms into the fault hypotheses. Then a knowledge base with rules that connect current network information to future faults can be established and updated automatically. By the knowledge base, we can get more hidden valuable information from the network management data and know more about the behavior of computer networks.
出处
《模式识别与人工智能》
EI
CSCD
北大核心
2000年第2期142-145,共4页
Pattern Recognition and Artificial Intelligence
基金
国家自然科学基金
