基于非法入侵检测的舰船网络安全预警系统  

Ship network security early warning system based on illegal intrusion detection

作　　者：蔡永强 CAI Yong-qiang(Wuhan University of Technology,Wuhan 430070,China;Guangxi Vocational College of Water Resources and Electric Power,Nanning 530023,China)

机构地区：[1]武汉理工大学,湖北武汉430070 [2]广西水利电力职业技术学院,广西南宁530023

出　　处：《舰船科学技术》2022年第9期158-161,共4页Ship Science and Technology

基　　金：广西职业教育教学改革项目(GXGZJG2021B104)。

摘　　要：以保护舰船网络安全为目的,设计基于非法入侵检测的舰船网络安全预警系统。系统在数据层采用Netflow V5采集舰船网络数据流,然后通过无线移动网络将其运输至容器层,通过Io T网关接收端连接网络服务器,实现交互同时封装、发布数据至中间层,中间层预处理舰船网络数据流并桥接环境与舰船,通过接口与应用层交互,在应用层通过预警模块解析、转化采集的舰船网络数据流,结合深度神经-记忆网络训练数据获取适宜非法入侵特征参数获取预警结果,经决策模块控制舰船主机后将非法入侵检测结果呈现在可视化界面,最后通过用户层实现整体过程管控。实验结果表明:该系统可有效控制网络会话连接延迟时间、抑制非法入侵行为,有效提升对于不同类型非法入侵的防御能力并完成预警,自动热启动累计次数跟随连续运行时间的增加而缓慢增加、具有较高的可靠性。In order to protect ship network security,a ship network security early warning system based on illegal intrusion detection is designed.System used in the data layer NetflowV5 ship network data flow,and then through the wireless mobile network,its transport to the container layer through IoT gateway receiving end connection network server,realize the interaction and encapsulation,release data to the middle tier,middle tier network data stream and preprocessing ships bridge environment and ships,interact through interfaces and application layer,In conversion of application layer through parsing,early warning module acquisition vessel network data flow,combined with the depth of neural network training-memory data acquisition for the illegal invasion of characteristic parameters for early warning results,after the decision module control ship host rendering the illegal intrusion detection result on the visual interface,and finally through the user layer to realize the whole process control.The experimental results show that the system can effectively control the network session connection delay time,suppress the illegal intrusion behavior,effectively improve the defense ability against different types of illegal intrusion and complete the early warning,the cumulative number of automatic hot start slowly increases with the increase of continuous running time,and has high reliability.

关 键 词：非法入侵检测 舰船网络 安全预警 无线移动网络 可视化 神经-记忆网络 

分 类 号：TN914[电子电信—通信与信息系统]