大型光纤通信网络断点检测模型仿真分析

发布时间：2018-09-13 13:11

【摘要】：对光纤网络断点进行准确检测,对维护网络安全是十分重要的。大规模光纤通信网络节点分布的分散,产生的断点存在随机性,单个断点也不会直接影响网络的通信。传统依据自主感应分析方法的断点检测模型在网络结构下,无法通过节点的通断来判断断点,需要逐个遍历网络节点进行关联,存在检测效率过低的问题。提出了一种依据免疫识别的神经网络光纤通信网络断点检测模型,按照免疫识别原理塑造神经网络检测器,通过训练将大型光纤通信网络的故障模式信息存储在分布检测器中,检测器用于采集光线通信网络的异常模式特征,当检测时应同特征样本匹配时则激活该检测器,按照检测器的激活状态检测大型光纤通信网络的断点,并给出了相应的训练方法。仿真结果表明,所提模型可准确检测出大型光纤通信网络断点,并且具有较高的检测效率。
[Abstract]:It is very important to detect the breakpoints of optical fiber network accurately for the maintenance of network security. Because of the dispersion of nodes in large scale optical fiber communication networks, the resulting breakpoints are random, and a single breakpoint will not directly affect the communication of the network. Under the network structure the traditional breakpoint detection model based on the autonomous induction analysis method can not judge the breakpoint through the on-off of the node. It needs to traverse the network nodes one by one to correlate with each other. Therefore the detection efficiency is too low. A neural network breakpoint detection model based on immune recognition is proposed. The neural network detector is modeled according to the principle of immune recognition, and the fault mode information of large optical fiber communication network is stored in the distributed detector by training. The detector is used to collect the abnormal mode features of the optical communication network. When the detection should match the characteristic sample, the detector is activated, and the breakpoints of the large optical fiber communication network are detected according to the active state of the detector. The corresponding training methods are also given. The simulation results show that the proposed model can accurately detect the breakpoints of large optical fiber communication networks and has a high detection efficiency.
【作者单位】：北京科技大学自动化学院;北京军区通信网络技术管理中心;国防大学信息指挥与作战教研部;
【基金】：国家自然科学基金(90818025);国家自然科学基金(61175122)
【分类号】：TN929.11;TN915.08

【参考文献】