基于数据驱动的电力光纤通信网脆弱节点动态挖掘

Data-driven Dynamic Mining of Vulnerable Nodes in Power Optical Fiber Communication Network

电力光纤通信网络脆弱节点受到外界影响或系统干扰时,会导致线路元件的相继故障,缩减网络的使用寿命,为此提出基于数据驱动的电力光纤通信网络脆弱节点动态挖掘方法。首先在大数据环境下剔除网络中的冗余节点,并根据剔除结果交换节点信息,确定通信网络脆弱节点位置连接后,获取通信网络的异常范围;再依据划分出的通信异常范围提取该区域中的节点脆弱性能特征,建立脆弱节点挖掘模型;最后依据该模型挖掘出通信网络中的脆弱节点,实现电力光纤通信网络脆弱节点动态挖掘。实验结果表明,利用该方法开展通信网络脆弱节点挖掘时,节点挖掘精度平均值为98%,网络传输时延最小值为19 ms,节点挖掘消耗时间为0.42 s,节点覆盖程度为96.88%,通信网络脆弱节点挖掘性能较好。

When vulnerable nodes in power fiber optic communication networks are affected by external factors or system interference,it can lead to the successive failures of line components,reducing the service life of the network.Therefore,a data-driven dynamic mining method for vulnerable nodes in power fiber optic communication networks is proposed.Firstly,in the big data environment,redundant nodes in the network are eliminated,and node information is exchanged based on the elimination results.After determining the location and connection of vulnerable nodes in the communication network,the abnormal range of the communication network is obtained.Then,the vulnerable performance characteristics of nodes in the area based on the divided communication anomaly range are extracted,and a vulnerable node mining model is established.Finally,based on this model,vulnerable nodes in the communication network are excavated,and dynamic mining of vulnerable nodes in the power fiber optic communication network is achieved.The experimental results show that when using this method to mine vulnerable nodes in communication networks,the average accuracy of node mining is 98%,the minimum network transmission delay is 19 ms,the consumption time of node mining is 0.42 s,and the node coverage is 96.88%.The mining performance of vulnerable nodes in communication networks is good.