自动检测技术在通信网络异常数据检测中的应用  

Application of Automatic Detection Technology in Anomaly Data Detection of Communication Networks

作　　者：肖媛 XIAO Yuan(Network Maintenance Center,Information Technology Branch of Shanxi Lu′an Chemical Group,Changzhi,Shanxi 046204,China)

机构地区：[1]山西潞安化工集团信息科技分公司网络维护中心,山西长治046204

出　　处：《自动化应用》2024年第19期123-125,共3页Automation Application

摘　　要：提出了一种基于长短时记忆网络和自适应阈值更新策略的自动检测技术,旨在提升通信网络异常数据检测的精准性。该技术利用LSTM网络对时间序列数据进行建模和特征提取,通过异常分数映射实现异常事件的识别。同时,引入自适应阈值更新策略,动态调整异常判定阈值,以适应数据分布的变化。结果表明,提出的技术在准确率等评价指标上均优于基线方法,为构建安全可信的通信网络环境提供了有力的技术支撑。A new automatic detection technique based on long short-term memory network and adaptive threshold update strategy is proposed to improve the accuracy of abnormal data detection in communication networks.This technology utilizes LSTM network to model and extract features from time series data,and identifies abnormal events through anomaly score mapping.At the same time,an adaptive threshold update strategy is introduced to dynamically adjust the anomaly determination threshold to adapt to changes in data distribution.The results indicate that the proposed technology outperforms the baseline method in evaluation metrics such as accuracy,providing strong technical support for building a secure and trustworthy communication network environment.

关 键 词：长短时记忆 自适应 自动检测 通信网络 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]