摘要
常规的通信网络安全告警结构多设定为独立形式,安全告警的范围难以进行扩展,导致告警响应时间不断延长,为此提出对基于卷积神经网络的通信网络安全告警方法的设计与验证分析,根据当前测试需求及标准的变化,先进行基础安全告警机制的制定,采用多目标的方式,扩展通信网络实际的安全告警覆盖范围,同时设计多目标交叉网络安全告警结构,以此为基础,构建卷积神经网络通信网络告警模型,采用自适应锁定的方式来实现安全告警处理。最终的测试结果表明:对比于传统时空特征融合通信网络安全告警测试组、传统邻域搜索粒子群通信网络安全告警测试组,此次所设计的卷积神经网络通信网络安全告警测试组最终得出的告警响应时间被较好地控制在0.25s以下,说明在卷积神经网络技术的辅助下,当前所设计的告警效果更佳,针对性更强,具有实际的应用价值。
The conventional communication network security alarm structure is mostly set as independent forms,The scope of security alarms is difficult to expand,Results in a prolonged alarm response time,The design and verification analysis of communication network security alarm method based on convolutional neural network are proposed,Based on changes in current test requirements and standards,First,develop the basic security alarm mechanism,In a multi-objective approach,Expand the actual security alarm coverage of the communication network,At the same time,multi-target cross network security alarm structure,On the basis,Constructing the communication network alarm model of convolutional neural network,Adaptive locking method is adopted to realize the security alarm processing.Final test results show that:compared to the traditional space-time features fusion communication network security alarm test group,the traditional neighborhood search particle swarm communication network security warning test group,the design of the convolutional neural network communication network security warning test group finally get alarm response time is better controlled under 0.25s,that under the aid of convolutional neural network technology,the current designed alarm effect is better,more targeted,has practical application value.
作者
王宁宁
WANG Ningning(Zhengzhou University of Industrial Technology Institute of Information Technology,ZhengZhou 450000,China)
出处
《长江信息通信》
2024年第2期93-95,113,共4页
Changjiang Information & Communications
关键词
卷积神经
神经网络
异常识别
通信网络
安全告警
告警方法
convolutional neural
neural network
anomaly identification
communication network
security alarm
alarm method
