摘要
为提高光通信网络数据传输能力,基于深度信念网络研究光通信网络数据异常识别方法。先构建光通信网络数据传输模型,采用深度信念网络进行传输信道均衡控制,利用模糊多分类支持向量机提取数据异常特征,构建数据分类学习模型,实现对数据异常重构和关联规则挖掘,采用深度信念网络对光通信网络数据进行异常张量切片重组,用张量对多关系网络进行建模,实现对光通信网络数据异常识别。仿真结果表明,所提改进方法的能量开销仅为1.2 kJ,生命周期为55.75 h,且识别时间仅为1.0 ms,优于其余两种方法,具有更大的应用价值。
In order to improve the data transmission capability of optical communication network,the data anomaly identification method of optical communication network is studied based on deep belief network.Firstly,the data transmission model of optical communication network is constructed,the deep belief network is used for transmission channel equalization control,the fuzzy multi classification support vector machine is used to extract the data anomaly features,the data classification learning model is constructed,the data anomaly reconstruction and association rule mining are realized,and the deep belief network is used to reorganize the anomaly tensor slice of optical communication network data,The multi relational network is modeled with tensor to realize the data anomaly identification of optical communication network.The simulation results show that the energy cost of the proposed improved method is only 1.2 kJ,the life cycle is 55.75 h,and the recognition time is only 1.0 ms,which is better than the other two methods and has greater application value.
作者
赵伟
王文娟
ZHAO Wei;WANG Wenjuan(Office of International Cooperation and Exchanges,Chongqing University of Posts and Telecommunications,Chongqing 400065,China;Information and Communications Branch,Sate Grid Chongqing Electric Power Company,Chongqing 401121,China)
出处
《激光杂志》
CAS
北大核心
2023年第1期174-178,共5页
Laser Journal
基金
国网重庆市电力公司信息通信分公司项目(No.SGCQXT00 JSXX2000556)。
关键词
深度信念网络
光通信网络
数据
异常识别
deep belief network
optical communication network
data
identify anomalies
