无界DoS攻击下的IPv6网络多层协同防御模型构建  

Construction of IPv6 Network Multi-Layer Collaborative Defense Model Under Unbounded DoS Attack

作　　者：吴元树 林少晶 WU Yuanshu;LIN Shaojing(College of Information Engineering,Fujian Business University,Fuzhou 350012,China)

机构地区：[1]福建商学院信息工程学院,福州350012

出　　处：《重庆科技学院学报（自然科学版）》2024年第3期68-74,共7页Journal of Chongqing University of Science and Technology：Natural Sciences Edition

基　　金：福建省科技计划引导性项目“基于智能转译的IPv6升级改造技术研究及其应用”(2021H0031)。

摘　　要：由于无界DoS攻击具有高度的隐蔽性和复杂性,单层防御无法解决入侵问题,因此构建了无界DoS攻击下的IPv6网络多层协同防御模型。根据协同防御思想,构建IPv6网络多层协同防御模型;收集模块流量,利用DNN模型对IPv6网络流量异常情况进行检测;引入协同式HCF自学习算法,以实现IPv6网络防御信息共享与异常网络流量过滤;利用攻击受损关联度的强化学习算法进行IPv6网络防御。实验结果表明,该模型可在2 ms内恢复CPU;在遭受不同类型的DoS攻击后IPv6网络吞吐量始终高于200 KiB s。这说明该模型可有效抵御不同类型的DoS攻击,反应迅速,防御效果好。Due to the high concealment and complexity of unbounded DoS attack,single-layer defense can not solve the intrusion problem.Therefore,a multi-layer collaborative defense model of IPv6 network under unbounded DoS attack is constructed,which is based on the concept of collaborative defense.Module traffic is collected and DNN model is used to detect abnormal IPv6 network traffic situations.The collaborative HCF self-learning algorithm is introduced to realize IPv6 network defense information sharing and abnormal network traffic filtering.IPv6 network defense is carried out by using the reinforcement learning algorithm of attack damage correlation.The experimental results show that the model can recover the CPU in 2 ms.IPv6 network throughput after different types of DoS attacks is consistently higher than 200 KiB s.This shows that the model can effectively resist different types of DoS attacks,react quickly and have good defense effect.

关 键 词：无界DoS攻击 IPV6网络 协同防御 深度神经网络 攻击源追踪 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]