文献类型：期刊文献

英文题名：Intelligent failure localization and maintenance of network based on reliability

作者：Zheng, Qing[1];Shao, Fangming[1]

机构：[1]East China Univ Sci & Technol, Sch Math, Dept Informat & Comp Sci, Shanghai 200237, Peoples R China

年份：2023

卷号：79

期号：1

起止页码：389

外文期刊名：JOURNAL OF SUPERCOMPUTING

收录：;EI(收录号：20222812354808);WOS:【SCI-EXPANDED(收录号:WOS:000825084700001)】；

语种：英文

外文关键词：Importance indicator; Monte Carlo algorithm; Failure localization; Back propagation neural network; Machine learning

摘要：The network maintenance and failure localization are particularly important to ensure the high reliability of network operation. In this work, we propose one importance indicator based on reliability theory to measure the importance of each link and design the corresponding Monte Carlo algorithm. Meanwhile, we adopt intelligent BP neural network to deal with the problem of failure localization by predicting the failure probability of each link. Simulations indicate that the approach can achieve a high localization accuracy and reduce the use of monitoring equipment effectively. Whether the failed link located is worth maintaining is determined by its importance indicator. The proposed approach can be used by service providers to reduce the cost on network failure localization and maintenance as well as maintain the high reliable operation of network. As the approach is not restricted to specific network technologies, it can be widely applied to different network types.