摘要
攻击入侵快速告警是全固态中波发射端监控系统安全防护体系中的重要组成部分,可以为系统安全防御决策提供重要依据。但当前告警水平比较低,实际中错误告警比例比较高,且系统攻击入侵告警时延比较长,为此文中提出了全固态中波发射端监控系统攻击入侵快速告警方法。该方法采用主成分分析法来筛选监控系统攻击入侵的特征,利用深度强化学习技术来检测监控系统攻击入侵行为,确定告警等级并作出相应的告警,完成对全固态中波发射端监控系统攻击入侵的快速告警。经实验证明,该方法应用下的监控系统攻击入侵错误告警比例能有效降低,告警时延也能有效缩短,具有良好的可行性与可靠性。
Attack intrusion alarm is an important part of the security protection system of all-solid-state medium-wave transmitter monitoring system,which can provide an important basis for system security defense decisions.However,the current alarm level is relatively low,the proportion of false alarms is relatively high in practice,and the system attack intrusion alarm delay is relatively long.Therefore,this paper proposes a rapid intrusion alarm method for all-solid-state medium-wave transmitter monitoring system.The method uses principal component analysis to screen the characteristics of monitoring system attack intrusion,and uses deep reinforcement learning technology to detect the intrusion behavior of monitoring system attack,determine the alarm level and make corresponding alarms to complete the rapid alarm of all-solid-state medium-wave transmitter monitoring system attack intrusion.Experiments show that the application of this method can effectively reduce the proportion of intrusion errors in the monitoring system,and the alarm delay can also be effectively shortened,which has good feasibility and reliability.
作者
苏成武
SU Chengwu(China Mobile Tietong Hefei Branch,Hefei 230011,China)
出处
《移动信息》
2024年第8期395-397,共3页
MOBILE INFORMATION
关键词
全固态中波发射端监控系统
攻击入侵
告警等级
主成分分析法
深度强化学习技术
All-solid medium wave transmitting terminal monitoring system
Attack intrusion
Alarm severity
Principal component analysis
Deep reinforcement learning technology