基于机器学习的通信网络安全故障检测方法研究
Research on fault detection methods for communication networks based on machine learning
王文燕 1王斌2
作者信息
- 1. 甘肃省军区,甘肃兰州 730000
- 2. 西部战区陆军参谋部,甘肃兰州 730000
- 折叠
摘要
[目的/意义]在大数据环境下网络通信具有数据资源海量化、网络状态动态化的特征.基于数据中心服务器、网络硬件的通信信息采集与传输,通常可能面临着网卡、网络参数配置和设备异常等故障发生.[方法/过程]通过对通信网络安全状态故障信息的采集,使用深度Q网络机器学习算法模型,对通信网络暂时故障、相似表征故障等信息做出诊断分析.[结果/结论]利用回合更新策略进行迭代训练,在降低信息采集开销的同时,提升通信网络故障的检测精度.
Abstract
[Purpose/Significance]In the big data environment,network communication has the characteristics of massive quantification of data resources and dynamic network status,and the collection and transmission of communication information based on data center servers and network hardware may usually face faults such as network card failures,network parameter configuration failures,and abnormal equipment failures.[Method/Process]Through the fault information collection for the security state of the communication network,the deep Q-network(DQN)machine learning algorithm model is used to diagnose and analyze the information such as temporary faults and faults with similar representations in the communication network.[Results/Conclusion]The round update strategy is used for iterative training,which not only reduces the overhead of information collection but also improves the accuracy of communication network fault detection.
关键词
机器学习/通信网络/故障检测/大数据/故障诊断/网络安全Key words
machine learning/communication network/fault detection/big data/fault diagnosis/network security引用本文复制引用
出版年
2024
NETL
NSTL
万方数据