摘要
在计算机网络系统中,由于存在外界攻击,设备寿命有限等原因,网络设备不可避免地会发生故障.本文提出了一种基于网络演算的故障检测方法,并证明了该方法能在确定的时延范围内检测出网络中节点发生故障及恢复正常的时间,以及节点的故障种类.在仿真算例上测试发现,本文的方法能及时检测出节点故障和正常的时间,且检测的平均准确率比当前常用的基于LSTM的方法高很多.
In computer networks, faults occurring to devices are common due to external interference or limited service life of devices. In this paper, we propose a fault detection method based on network calculus. We show that the method can detect the time of faults occurring or returning to normal and the type of fault in predetermined delay. Simulation show that our algorithm can detect fault or normal in time, and the accuracy is much higher than algorithm based on LSTM.
作者
章子游
赵千川
杨文
ZHANG Zi-you;ZHAO Qian-chuan;YANG Wen(Department of Automation and BNRist,Tsinghua University,Beijing 100084,China;Key Laboratory of Space Launching Site Reliability Technology,Haikou Hainan 570100,China)
出处
《控制理论与应用》
EI
CAS
CSCD
北大核心
2019年第11期1861-1870,共10页
Control Theory & Applications
基金
国家“十三五”重点研发项目(2017YFC0704100,2016YFB0901900)
国家自然科学基金项目(61425027)
“111”引智计划(BP2018006)
北京国家信息科学与技术研究中心项目(BNR2019TD01009)资助~~
关键词
计算机网络
故障检测
网络演算
时延
computer networks
fault detection
network calculus
delay
